Loading [MathJax]/jax/output/SVG/jax.js
Search Advanced

Mathematical Biosciences and Engineering

2023, Volume 20, Issue 4: 6630-6651. doi: 10.3934/mbe.2023285
Previous Article Next Article
Research article Special Issues

A-RetinaNet: A novel RetinaNet with an asymmetric attention fusion mechanism for dim and small drone detection in infrared images

  • College of Information Engineering, Shanghai Maritime University, Shanghai 201306, China

  • Academic Editor: Hamidreza Namazi

  • To solve the problems of texture lacking and resolution coarseness in the detection of dim and small drone targets in infrared images, we propose a novel RetinaNet with an asymmetric attention fusion mechanism for dim and small drone detection. First, we propose a super-resolution texture-enhancement network as an effective solution for the lack of texture-related information on small infrared targets. The network generates super-resolution images and enhances the texture features of the targets. Second, considering the inadequacy of feature pyramids in the feature fusion stage, we use an asymmetric attention fusion mechanism to constitute an asymmetric attention fusion pyramid network for cross-layer feature fusion in a bidirectional manner; it achieves high-quality semantic and location detail information interaction between scale features. Third, a global average pooling layer is employed to capture global spatial-sensitive information, thus effectively identifying features and achieving classification. Experiments were conducted by using a publicly available infrared image dim-small drone target detection dataset; the results show that the proposed method achieves an AP of 95.43% and a recall of 80.6%, which is a significant improvement over the current mainstream target detection algorithms.

    Citation: Zhijing Xu, Jingjing Su, Kan Huang. A-RetinaNet: A novel RetinaNet with an asymmetric attention fusion mechanism for dim and small drone detection in infrared images[J]. Mathematical Biosciences and Engineering, 2023, 20(4): 6630-6651. doi: 10.3934/mbe.2023285

    Related Papers:

    [1] Xiran Yang, Yong Qi . Study on the evolution driving mechanism of IPM to regional economic development. Mathematical Biosciences and Engineering, 2022, 19(7): 6699-6711. doi: 10.3934/mbe.2022315
    [2] Yanhong Feng, Qingqing Hu . Heterogeneity and spillover effects of carbon emission trading on green innovation. Mathematical Biosciences and Engineering, 2023, 20(4): 6468-6497. doi: 10.3934/mbe.2023279
    [3] Jianquan Guo, Guanlan Wang, Mitsuo Gen . Green closed-loop supply chain optimization strategy considering CER and incentive-compatibility theory under uncertainty. Mathematical Biosciences and Engineering, 2022, 19(9): 9520-9549. doi: 10.3934/mbe.2022443
    [4] Chrysovalantis Malesios, Nikoleta Jones, Alfie Begley, James McGinlay . Methodological approaches to exploring the spatial variation in social impacts of protected areas: An intercomparison of Bayesian regression modeling approaches and potential implications. Mathematical Biosciences and Engineering, 2024, 21(3): 3816-3837. doi: 10.3934/mbe.2024170
    [5] Eduardo Gonzalez-Olivares, Alejandro Rojas-Palma . Global stability in a modified Leslie-Gower type predation model assuming mutual interference among generalist predators. Mathematical Biosciences and Engineering, 2020, 17(6): 7708-7731. doi: 10.3934/mbe.2020392
    [6] Nariyuki Nakagiri, Hiroki Yokoi, Yukio Sakisaka, Kei-ichi Tainaka . Population persistence under two conservation measures: Paradox of habitat protection in a patchy environment. Mathematical Biosciences and Engineering, 2022, 19(9): 9244-9257. doi: 10.3934/mbe.2022429
    [7] Yanjun Liu, Chin-Chen Chang, Peng-Cheng Huang . Security protection using two different image shadows with authentication. Mathematical Biosciences and Engineering, 2019, 16(4): 1914-1932. doi: 10.3934/mbe.2019093
    [8] Yuta Okada, Taishi Kayano, Asami Anzai, Tong Zhang, Hiroshi Nishiura . Protection against SARS-CoV-2 BA.4 and BA.5 subvariants via vaccination and natural infection: A modeling study. Mathematical Biosciences and Engineering, 2023, 20(2): 2530-2543. doi: 10.3934/mbe.2023118
    [9] Tinghui Li, Jieying Wen, Danwei Zeng, Ke Liu . Has enterprise digital transformation improved the efficiency of enterprise technological innovation? A case study on Chinese listed companies. Mathematical Biosciences and Engineering, 2022, 19(12): 12632-12654. doi: 10.3934/mbe.2022590
    [10] Luis E. Ayala-Hernández, Gabriela Rosales-Muñoz, Armando Gallegos, María L. Miranda-Beltrán, Jorge E. Macías-Díaz . On a deterministic mathematical model which efficiently predicts the protective effect of a plant extract mixture in cirrhotic rats. Mathematical Biosciences and Engineering, 2024, 21(1): 237-252. doi: 10.3934/mbe.2024011

  • To solve the problems of texture lacking and resolution coarseness in the detection of dim and small drone targets in infrared images, we propose a novel RetinaNet with an asymmetric attention fusion mechanism for dim and small drone detection. First, we propose a super-resolution texture-enhancement network as an effective solution for the lack of texture-related information on small infrared targets. The network generates super-resolution images and enhances the texture features of the targets. Second, considering the inadequacy of feature pyramids in the feature fusion stage, we use an asymmetric attention fusion mechanism to constitute an asymmetric attention fusion pyramid network for cross-layer feature fusion in a bidirectional manner; it achieves high-quality semantic and location detail information interaction between scale features. Third, a global average pooling layer is employed to capture global spatial-sensitive information, thus effectively identifying features and achieving classification. Experiments were conducted by using a publicly available infrared image dim-small drone target detection dataset; the results show that the proposed method achieves an AP of 95.43% and a recall of 80.6%, which is a significant improvement over the current mainstream target detection algorithms.


    1.   Introduction

    For decades, China has relied too much on high-polluting and high-energy-consuming industries to drive its economy. This development model, which over emphasized economic growth while neglecting environmental protection, although boosting the country's economy in the short run, has seriously damaged China's ecological environment. As the environmental situation continues to deteriorate, China's previous fast rate of innovation, which primarily relied on the direct consumption of raw materials, needs to transform to high-quality innovation-characterized by green, intelligent, and sustainable development. In 2015, the Fifth Plenary Session of the 18th Central Committee proposed the Five Development Concepts, in which "green" and "innovation" were included as two important concepts, indicating that China's attention to green innovation had reached a new height. In 2021, the 14th Five-Year Plan proposed the implementation of a green technology innovation campaign, insisting on the green development concept. Enterprises play a significant role in boosting economic growth, but they also consume a considerable amount of natural resources. Promoting enterprise green transformation and accelerating the development of enterprise green technology innovation not only constitute an important measure to promote coordinated economic growth and environmental protection, they also constitute an effective path to achieve both economic efficiency and ecological protection. While R&D investment has continued to increase in China, regional green innovation performance has not increased proportionately [1].

    Promoting enterprise green technology innovation cannot be achieved without the intellectual property rights (IPR) protection policies. Although scholars like Franklin Allen pointed out that the case of China contradicts the findings in the literature on law, institutions, finance, and growth, because, despite having underdeveloped legal and financial systems, it has managed to achieve one of the most rapid economic growth rates [2]; furthermore, sufficient evidence proved that a progressive legal system and technological advancements, supported by IPR policies have become a crucial factor in China's overall national strength competition [3,4,5]. Since the release of the National Innovation-Driven Development Strategy Outline, China's scientific and technological innovation has achieved rapid development with strong support and promotion from governments at all levels. In accordance with the 2019 Global Innovation Report, China's innovation ranking has risen to the 14th in the world.

    However, China's IPR infringement issues are still not trending positive, and the IPR protection system urgently needs improvement. The results of the Business Confidence Survey showed that 78% of the surveyed companies had encountered IPR issues. The severe situation of IPR protection has made people's demand for improving the IPR protection system increasingly urgent. The Outline for Building Intellectual Property Power (2021–2035) emphasizes the importance of IPR protection. IPR protection provides strong institutional guarantees for scientists and enterprises to coordinate innovation in order to truly realize the transformation of applied basic research and new technologies and other achievements, thus promoting global sustainable and green technological innovation.

    With the profound understanding of scholars on IPR protection and the deepening study on green innovation development, a large amount of related research has gradually emerged, mainly focusing on the factors affecting green innovation development. Firstly, there exists an extensive body of research that examines how environmental regulations affect the advancement of eco-friendly innovation in enterprises. Certain academics have concluded that environmental regulations that are market-based can offer stronger incentives for innovation [6,7]. According to certain studies, the implementation of low-carbon city pilot policies can drive innovation in green technologies related to energy conservation and alternative energy in enterprises [8,9]. Many Chinese scholars believe that direct environmental regulations have a more significant incentive effect on enterprises with strong innovation capabilities, and it is necessary for the Chinese government to guide enterprises through market mechanisms rather than using a one-size-fits-all approach [10,11].

    Second, the effects of financial development on green innovation are studied in depth. Scholars generally believe that from the market's perspective, developing green finance can promote the innovation of green technology in enterprises. Some scholars have pointed out that the development of green finance by means of green credit can optimize the economic structure, thus creating incentives for enterprises to innovate green technologies and helping them to make the green transition [12,13]. Specifically, the stock market is found the most strongly associated with green technological innovation, and the better the development of the stock market, the stronger the promotional effect on green technological innovation [14,15]. Li et al. [16] found that green credit and related environmental protection subsidies can encourage enterprises to carry out clean production. Furthermore, certain academics have examined the correlation between green finance and eco-friendly innovation in countries with varying levels of development. Their findings suggest that green finance has a positive impact on green innovation in emerging countries, but it can have a negative impact in countries that have already achieved significant progress in green innovation [17,18].

    Moreover, the effect of IPR protection on innovation has been a focus of academic study in the recent decade. Dinopoulos and Segerstrom [19] argued that IPR protection could promote regional innovation by increasing the rate of international technology transfer within multinational firms. Hudson and Minea [20] opined that the impact of IPR on innovation is intricate and can exhibit nonlinear trends, which are dependent on the initial levels of both IPR and per capita GDP. With a world sample, Sweet & Maggio [21] found that stronger IPR protection could promote innovation advances only in countries with above-average levels of development and complexity. Certain scholars from China have identified an inverted U-shaped connection between IPR protection and the innovation capability of industrial enterprises in China. However, most of the samples analyzed were located on the lower side of the turning point, suggesting that reinforcing IPR protection could still prove advantageous in enhancing China's industrial innovation [22,23]. Papageorgiadis and Sharma [24] focused on the relationship between innovation and the strength of IPR regulations and enforcement, and the researchers discovered that the level of IPR regulations and enforcement has a noteworthy impact on promoting innovation. Grimaldi et al. [25] found that not having any IPR protection strategy can be a barrier to outbound green innovation.

    Prior theoretical and empirical work have provided a solid foundation and valuable reference for our study, but there is still room for further research, especially in emerging countries like China. In this study, we focus on the impact of IPR protection on green innovation development in China, with the marginal contribution mainly being as follows. First, the study takes the correlation between IPR protection and enterprise environmentally friendly innovation in China as its research object. Existing research mainly centers on the effect of IPR protection on innovation in general, with little research focusing on green innovation development. This paper regards the pilot policy of IPR protection as an external shock, exploring the effect of external regulations on the internal development of corporate green innovation. Second, the research delves into how IPR protection affects environmentally friendly innovation in Chinese enterprises, particularly exploring the mediating role of R&D investment and financing constraints in the impact of IPR pilot protection policy on the development of green technology innovation.

    The rest of this paper is structured as follows. In Section 2, the research scheme is presented, which includes proposing research hypotheses based on theoretical analysis and constructing research models. Section 3 specifies the indicator selection and data sources. Section 4 discusses the empirical test on the impact of IPR on enterprise green innovation. Section 5 furthers the analysis on the impact mechanism. Section 6 concludes the study and provides policy implications.

    2.   Research scheme

    2.1.   Theoretical analysis and research hypotheses

    2.1.1.   Effects of IPR protection on enterprise green innovation

    Justification of the exclusiveness of an intellectual property right involves more than one theory, and typical theories involved include the theory of incentives for creation and the theory of compensation for public opening [26,27,28,29,30]. Therefore, strengthening judicial protection of intellectual property is a fundamental guarantee for the promotion of the vitality and driving force of enterprise innovation. Compared to traditional innovation, green innovation has a double externality issue, which results in a sub-optimal investment in green innovations; thus, policy regulations play a determinant role in motivating eco-friendly innovation [31,32,33] and thus reduce the incentives for its adoption. IPR protection can have a positive impact on enterprise green innovation by increasing the value of R&D results in environmental protection. In other words, when the protection of intellectual property is insufficient, the green innovation outcomes of enterprises will be imitated by competitors at will, resulting in lower-than-expected returns on green innovation and reduced profits, which will weaken the incentive for companies to pursue environmentally friendly innovation [34,35,36]. Therefore, the first hypothesis of this study is proposed.

    Hypothesis 1: IPR protection promotes the development of enterprise green innovation.

    2.1.2.   Analysis of the mediating effect of R&D investment in the impact of IPR protection on enterprise green innovation

    R&D investment is one of the decisive factors in determining the output of enterprise green innovation. Studies have shown a correlation between the R&D expenses of 28 E.U. nations and their ecological footprints [37]. It profoundly affects the entire process of enterprise innovation and is the economic foundation of the enterprise green technology innovation capabilities [38]. Some research has shown that increasing enterprise R&D investment can ensure their green technology improvement and innovation levels, guaranteeing their market competitiveness and good performance [39]. IPR protection can increase the enterprise R&D investment through the crowding-out effect. IPR protection can increase the imitation cost and infringement cost among competing enterprises. The increasing costs in introducing external technologies force enterprises to increase their R&D investment to obtain market competitiveness, which promotes green technology innovation [40,41]. Therefore, we propose the second hypothesis.

    Hypothesis 2: R&D investment is a channel through which IPR protection promotes enterprise green innovation.

    2.1.3.   Analysis of the mediating effect of enterprise financing constraints in the impact of IPR protection on enterprise green innovation

    The ability of enterprises to obtain sufficient external financing is one of the main factors affecting the success of R&D and innovation activities [42,43]. In addition, compared with traditional technological innovation, green innovation requires significant upfront investment and has a longer payback period. Therefore, it requires a certain amount of financial support to address a series of market mechanism problems that may arise during the innovation process [44,45,46]. This means that financing constraints are an important limiting factor in the process of enterprise green innovation. R&D and innovation activities of enterprises have information asymmetry characteristics. When external investors are not aware of the enterprise's green innovation achievements, they cannot evaluate the commercial value of this technology. However, when they are fully aware, they may directly imitate the enterprise's innovation achievements, making the enterprise unwilling to fully disclose information related to its R&D and innovation activities [47,48]. Therefore, effective solutions to the information asymmetry between external investors and enterprises constitute the reason why IPR protection can alleviate enterprise financing constraints. IPR protection can effectively restrict potential imitation and infringement behaviors, consequently prompting enterprises to provide necessary intellectual property and technology information to external investors, reducing information asymmetry between external investors and enterprises, helping enterprises to accelerate financing speed and reducing financing costs, thus providing a good foundation for enterprise green technology innovation activities [49]. Thus, the third hypothesis is proposed.

    Hypothesis 3: Alleviating enterprise financing constraints is the channel through which IPR protection promotes enterprise green innovation.

    2.2.   Model setting

    2.2.1.   Construction of the difference-in-differences model

    The difference-in-differences (DID) model is commonly used in econometrics to quantitatively evaluate the implementation effects of public regulations or projects. The DID model uses data obtained from quasi-natural experiments based on non-randomly assigned regulation experimental and control groups. By controlling for preexisting differences between the research subjects, the model effectively separates the actual effects of policy impacts. In 2014, Beijing, Shanghai, and Guangzhou were chosen as pilot cities to establish intellectual property courts, while no such court had previously been established in China. This meets the prerequisite conditions required by the DID model. To eliminate unobserved impacts on the results, we constructed a DID model with fixed years and enterprises to evaluate the impact of IPR protection on the development of enterprise green innovation. The model setting was as follows:

    lngreenpatenti,t=β0+β1policyi,t+βjcontrolsi,t+year+firm+εi,t (1)

    where i stands for the enterprise, t represents the year, and εi,t is the random error term; lngreenpatent represents the explained variable, i.e., green innovation, and policy=treat∗post is the core explanatory variable, i.e., the pilot policy of the intellectual property court; treat represents the treatment effect, where enterprises located in Beijing, Shanghai, and Guangzhou are defined as the experimental group and assigned a value of 1, while others are defined as the control group and assigned a value of 0; post stands for the time effect, where years after 2015 (the year the pilot policy put into effect) are assigned a value of 1, while other years are assigned a value of 0; the company asset size (size), company age (age), company leverage ratio (LEV), company total asset turnover ratio (ROA), number of board members (board), percentage of independent directors (Indep), nature of company's ownership (SOE), and company annual revenue growth rate (growth) are control variables.

    2.2.2.   Building a mediating effect model

    To analyze whether the R&D investment and the financing constraints play a mediating role in promoting green innovation development through IPR protection, we have introduced a mediating variable to construct a mediating model. Following the practices of Sun and Song [50] as well as Huang et al. [51], we adopted the per capita R&D investment of enterprises as a measure of the level of enterprise R&D investment. In reference to the practices of Hadlock and Pierce [52] and Song et al. [53], this study adopts the FC (financial constraints) indicator in the SA (size-age) index as a measure of financial constraints. Both the per capita R&D investment and the FC indicator were taken as mediating variables to establish a stepwise regression model for testing the mediating effect [54,55]. The results were further validated by using the Sobel test. The specific models are as follows:

    (i) With the per capita R&D investment as the mediating variable:

    greenpatenti,t=β0+β1policyi,t+βjcontrolsi,t+year+firm+εi,t (2)
    R&Di,t=β0+β1policyi,t+βjcontrolsi,t+year+firm+εi,t (3)
    greenpatenti,t=β0+β1policyi,t+β2R&Di,t+βjcontrolsi,t+year+firm+εi,t (4)

    (ii) With the FC indicator as the mediating variable:

    greenpatenti,t=β0+β1policyi,t+βjcontrolsi,t+year+firm+εi,t (5)
    FCi,t=β0+β1policyi,t+βjcontrolsi,t+year+firm+εi,t (6)
    greenpatenti,t=β0+β1policyi,t+β2FCi,t+βjcontrolsi,t+year+firm+εi,t (7)

    where R&D represents the mediating variable for enterprise R&D investment, FCrepresents the mediating variable for financial constraints, and the variables have the same meanings as the ones in Eq (1).

    3.   Indicator selection and data sources

    3.1.   Indicator selection

    The explained variable in this study is the level of green innovation of listed companies. According to the division of green patents in the International Patent Classification (IPC) Green Inventory developed by the IPC Committee of Experts, green patents are divided into two categories: green invention and green utility model patents. Following the approach of Deng et al. [56], this research measures the level of green innovation by using the application number of green invention patents as the indicator. This choice is based on two reasons: one is that the application threshold for green invention patents is higher than that for green utility model patents, and that there is a greater requirement for the company's green innovation level [57]; the other is that due to the complex and long patent authorization process, relative to the number of patent authorizations, the number of patent applications is believed to have the features of timeliness, reliability, and stability [58]. Since the application number of green invention patents has a right-skewed distribution feature, this study adopts the number of green innovation patent applications plus one taking the natural logarithm to measure the explained variable.

    The core explanatory variable is the IPR protection pilot policy, which is measured by the interaction between the location and the time of the intellectual property court pilot policy implementation. The reason for using the interaction term as the core explanatory variable is that, since the research objective was to evaluate the impact of IPR protection policies on the level of green innovation of enterprises, it is necessary to examine the policy change-related factors, which involve time and place, that is, the time and space factors. Spatial factors and the economic status of Chinese cities have a close correlation; the high economic status of the city is often the pilot area of the policy, and mainland China's policy pilots tend to start from the oversized first-tier cities. In terms of the policy implementation location variable, enterprises with offices located in the three pilot cities were defined as the treatment group and assigned a value of 1, while the rest were defined as the control group and taken a value of 0. Regarding the policy implementation time variable, the years after 2015 had a value of 1, and other years had a value of 0.

    To control the impact of other economic indicators on green innovation in enterprises, we have introduced the following control variables based on existing literature: 1) Enterprise size, which is calculated by the year-end total assets. Generally, the size of a company has a significant effect on its innovation. Larger companies tend to raise their R&D investment to ensure their technology development level and competitiveness. 2) Enterprise age, which represents the maturity level of an enterprise. Studies have shown that the more mature an enterprise, the stronger its innovation consciousness [59,60]. This study measures the enterprise age by the length of time it has been listed. 3) Enterprise debt. Reasonable debt management can provide enterprises with more capital for technology and research activities. This study uses the asset-liability ratio of the enterprise as an indicator of enterprise debt. 4) Enterprise profitability, which is calculated by the return on assets (ROA) in this study. The higher the ROA, the stronger the company's profitability. It is generally believed that companies with higher profitability have stronger innovation capabilities. 5) Enterprise management level, which is evaluated by the number of board members and the percentage of independent directors, because the structure of the director board and the percentage of independent directors reflect the decision-making level and governance structure of the company's management. 6) Enterprise ownership. Different enterprise ownership leads to different levels of technological innovation. If an enterprise is a state-owned enterprise (SOE), the value is 1, and if it is a non-SOE, the value is 0. 7) Enterprise development capability, which is assessed by the annual revenue growth rate, reflecting the company's development status. Generally, the greater the company's development potential, the more it will increase its investment in technological innovation.

    The variables selected and specific measurement indicators for the empirical study in this paper are presented in Table 1.

    Table 1.  Variables and measurement.
    Variable type Symbol Indicator Measurement
    Explained variable green patent Green innovation Number of green invention patent applications
    Explanatory variable policy IPR protection policy Pilot location × Pilot time
    Control variables size Enterprise size Year-end total assets of the enterprise
    age Enterprise age The length of time a company has been listed
    LEV Enterprise debt Asset-liability ratio of the enterprise
    ROA Enterprise profitability ROA of the enterprise
    board Enterprise management level Number of directors in the corporate board
    Indep Proportion of independent directors in the enterprise board
    SOE Enterprise ownership SOE/non-SOE (1/0)
    growth Enterprise development capability Annual revenue growth rate
     | Show Table
    DownLoad: CSV

    3.2.   Data source

    Since the intellectual property pilot courts (put into effect in 2015) are relatively new, to ensure the symmetry of data before and after the establishment of the intellectual property pilot courts and the comparability of results, we selected green invention patent application data and the corresponding economic data of A-share listed companies in China from 2011 to 20191. The green patent data of listed companies were obtained from the Chinese National Intellectual Property Administration, and the selection of green patent applications is based on the International Patent Classification (IPC) Green Inventory. The relevant economic data of companies are acquired from the China Stock Market & Accounting Research (CSMAR) Database.

    1 To prevent shocks to the results from the financial crisis (2008 global financial crisis) and the major public health crisis (COVID-19 epidemic), the data sample in this paper does not include the time periods of 2008–2010 and 2020 to the present.

    To ensure data validity, we processed the original samples of listed companies as follows: 1) considering that the data of financial enterprises are not suitable for measuring the level of enterprise green innovation, companies belonging to the financial industry are excluded; 2) companies without continuous five-year observations or are under special treatment (ST) were excluded; 3) to avoid the influence of extreme values, all continuous variables selected were winsorized at the upper and lower 1%. Finally, a valid annual sample of 19,285 companies was obtained.

    The descriptive statistics of all variables involved in the empirical analysis are shown in Table 2. The specific statistical characteristics of the variables are as follows:

    Table 2.  Variables and descriptive statistical results.
    Variable symbol Sample number Mean Standard deviation Minimum Maximum
    green patent 19,285 3.953 23.309 0 915
    policy 19,285 0.113 0.316 0 1
    size 19,285 22.219 1.247 19.885 25.93
    age 19,285 2.829 0.347 1.792 3.466
    LEV 19,285 0.424 0.204 0.057 0.896
    ROA 19,285 0.039 0.058 -0.193 0.206
    board 19,285 2.138 0.195 1.609 2.639
    Indep 19,285 0.375 0.054 0.333 0.571
    SOE 19,285 0.378 0.485 0 1
    growth 19,285 0.164 0.376 -0.543 2.213
     | Show Table
    DownLoad: CSV

    It can be seen in Table 2 that the standard deviations of both the explanatory variable and the explained variable are relatively large, indicating significant heterogeneity in both the green technology innovation and local intellectual property protection policies of enterprises. By comparing the coefficients of variation between the explanatory variable and the explained variable, we found that the coefficient of variation of green technology innovation is larger, i.e., close to 600%, and that of the intellectual property policy change is close to 300%. As for the control variables, only the enterprise size has a relatively large coefficient of variation, but it is only 5.6%.

    4.   Econometric tests of the impact of IPR protection on enterprise green innovation

    4.1.   Benchmark regression

    This research adopted a DID model to assess the impact of IPR protection on enterprise green innovation, and the DID benchmark regression results are presented in Table 3, where Columns (1) and (2) control for fixed effects at the enterprise level, while Columns (3) and (4) control for fixed effects at both enterprise and time levels; additionally, control variables are added to Columns (2) and (4). All regression analyses were performed by using the enterprise-level clustering adjusted standard errors.

    Table 3.  Benchmark regression results.
    (1) (2) (3) (4)
    lngreen patent lngreen patent lngreen patent lngreen patent
    policy 0.410*** 0.0731** 0.0846** 0.0732**
    (0.0298) (0.0306) (0.0329) (0.0320)
    size 0.283*** 0.266***
    (0.0226) (0.0232)
    age 0.598*** 0.171
    (0.0532) (0.137)
    Lev -0.127** -0.0782
    (0.0642) (0.0649)
    ROA -0.0299 0.0276
    (0.124) (0.124)
    board 0.0270 0.0177
    (0.0707) (0.0702)
    Indep -0.0305 -0.0693
    (0.215) (0.214)
    SOE 0.166*** 0.169***
    (0.0557) (0.0556)
    growth -0.0198 -0.0203
    (0.0123) (0.0125)
    constant -7.448*** -5.955***
    (0.476) (0.648)
    controls NO YES NO YES
    year NO NO YES YES
    firm YES YES YES YES
    R-squared 0.024 0.150 0.124 0.154
    N 19,285 19,285 19,285 19,285
    Note: *, **, and *** represent the significance levels of 10%, 5%, and 1%, respectively. Standard errors of the regression coefficients are shown in parentheses.
     | Show Table
    DownLoad: CSV

    As shown in Table 3, it is obvious that IPR protection promotes enterprise green innovation. The establishment of intellectual property pilot courts effectively promotes the technological innovation of enterprise green technology, thus verifying Hypothesis 1. The regression coefficient relative to green invention patents in Column (1) is 0.410, which is significant at the level of 1%. When controlling for the year and company fixed effects, the regression coefficient of relative to green invention patents in Column (3) is 0.0846, significant at 5% and smaller than that in Column (1). This suggests that there are time and firm factors that affect green technology innovation in enterprises. Therefore, controlling for firm and time fixed effects effectively eliminates the interference of some random factors and enables a more accurate analysis of the net impact of the pilot policy implementation.

    Considering that various company characteristics, such as debt, management structure, and development capabilities, can also have an impact on green technology innovation, selected control variables were further added to the model. After adding control variables, the regression coefficients relative to green invention patents in Columns (2) and (4) were 0.0731 and 0.0732, respectively, significant at 5% and still supporting the expected result of this study.

    The results shown in Table 3 indicate that augmenting IPR protection is conducive to enhancing the development capability of green innovation in enterprises. However, most evaluation criteria prioritize quantity over quality, resulting in a large amount of low-quality innovation that pursues short-term benefits in the innovation process, which not only wastes resources, going against the concept of green development, but also fails to solve the bottleneck of high-end technology faced by China. Therefore, the Chinese government should vigorously promote the establishment and improvement of the IPR protection system, as represented by the intellectual property court, and strengthen the protection of green innovation rights and interests, creating an environment that truly conforms to the concept of green innovation.

    4.2.   Parallel trend test

    The prerequisite for adopting the DID method is that the trends of change between the treatment group and the control group must be consistent before being processed. Based on this, we decided to demonstrate the stability of the model through a parallel trend test. Figure 1 displays the parallel trends of enterprise green innovation output from 2011 to 2019. The solid line represents the average number of green patent applications in the treatment group, while the dashed line represents the average number of green patent applications in the control group. The vertical dashed line represents the start of the IPR pilot policy taking effect.

    Figure 1.  Parallel trends.
    DownLoad: Full-Size Img PowerPoint

    The DID model of this research passed the robustness test. As indicated in Figure 1, before the establishment of intellectual property pilot courts, the trend of the number of green invention patent applications in the experimental group (Beijing, Shanghai, Guangzhou) was similar to that in the control group, maintaining a parallel trend. However, after implementing the pilot policy, the gap between the two groups of green invention patents significantly widened. The number of green invention patents in the three pilot cities grew much faster than in other cities, and this trend was maintained until the end of the sample period. These results indicate that the DID model passed the parallel trend test. The specific results are presented in Table 4.

    Table 4.  Parallel trend test results.
    (1) (2)
    Variables lngreen patent lngreen patent
    d_3 0.0183 0.0183
    (0.0395) (0.0395)
    d_2 0.0178 0.0178
    (0.0409) (0.0409)
    d_1 0.0512 0.0512
    (0.0447) (0.0447)
    current 0.0542 0.0542
    (0.0462) (0.0462)
    d1 0.135*** 0.135***
    (0.0503) (0.0503)
    d2 0.122** 0.122**
    (0.0525) (0.0525)
    d3 0.0393 0.0393
    (0.0554) (0.0554)
    d4 0.133** 0.133**
    (0.0579) (0.0579)
    Note: *, **, and *** represent the significance levels of 10%, 5%, and 1%, respectively; the standard errors of the regression coefficients are shown in parentheses.
     | Show Table
    DownLoad: CSV

    Table 4 reveals that before the establishment of the intellectual property pilot courts, there was no significant difference between the experimental and control groups. However, in the "current" year (i.e., the first year after the policy was implemented), the difference between the two groups began to be significant, indicating that the pilot policy of intellectual property courts directly impacted the green technology innovation of enterprises in the experimental group. No lag effect was observed, which again confirms that the experimental and control groups passed the parallel trend test.

    5.   The impact mechanism of the IPR protection on enterprise green innovation

    5.1.   The mediating effect of the R&D investment

    We employed the stepwise method and the Sobel test to confirm the mediating effect of the R&D investment in the influence of the IPR protection on enterprise green innovation, and Table 5 shows the test results.

    Table 5.  The mediating effect of the R&D investment.
    Plan A: The mediating effect test results
    (1) (2) (3)
    Explanatory variables Total effect R&D investment Mediating effect of R&D investment
    Policy 0.0815** 0.376** 0.0748**
    (0.0369) (0.191) (0.0367)
    lnrd 0.0177***
    (0.00354)
    Size 0.325*** 0.691*** 0.313***
    (0.0274) (0.116) (0.0276)
    Age -0.0205 -0.604 -0.00985
    (0.151) (0.631) (0.150)
    Lev -0.135* -1.184*** -0.114
    (0.0775) (0.391) (0.0774)
    ROA -0.0625 0.922 -0.0788
    (0.141) (0.721) (0.140)
    Board 0.0378 -0.324 0.0435
    (0.0770) (0.353) (0.0771)
    Indep -0.0448 -0.640 -0.0335
    (0.236) (0.943) (0.234)
    SOE 0.147** 0.119 0.145**
    (0.0643) (0.310) (0.0634)
    Growth -0.0204 -0.138** -0.0180
    (0.0154) (0.0661) (0.0155)
    Constant -6.708*** -9.795*** -6.535***
    (0.738) (2.855) (0.740)
    Controls YES YES YES
    Year YES YES YES
    Firm YES YES YES
    Observations 16,278 16,278 16,278
    Number of stkcd 2,147 2,147 2,147
    R-squared 0.167 0.188 0.170
    Plan B:Mediating effect test
    Mediating variable c a σa b σb Z Mediating effect weight
    R&D investment 0.2223 2.2792 0.0931 0.0566 0.0018 19.1*** 58.06%
    Note that *, **, and *** indicate significant levels of 10%, 5%, and 1%, respectively; the standard errors of the regression coefficients are shown in parentheses.
     | Show Table
    DownLoad: CSV

    The results in Table 5 show that R&D partially mediates the impact of IPR protection on promoting enterprise green innovation. In Column (1), the regression coefficient for policy is 0.0815, which is significant at the 5% level, indicating that IPR protection has a positive promotional effect on enterprise green technology innovation. Therefore, further testing was performed. Column (2) presents the regression estimation for IPR and R&D, with a regression coefficient for policy of 0.376, which is also significant at 5%, indicating that the protection of intellectual property exerts a positive driving impact on increasing R&D investment in enterprises. Column (3) outlines whether the level of enterprise innovation in green technology will be greatly impacted by the presence of IPR protection and R&D investment. The results show that IPR protection still significantly affects green innovation in enterprises at the 5% level and that the regression coefficient for R&D investment is significant at the 1% level. This indicates that there is a partial mediating effect between IPR protection and enterprise green innovation. Strengthening IPR protection can increase enterprise R&D investment and thereby have a positive effect on green technology innovation. R&D investment is the transmission path through which IPR protection affects green technology innovation in enterprises, thus Hypothesis 2 is verified.

    In order to confirm that R&D investment has a crucial role in the impact of IPR protection on the innovation of green technology in enterprises, we further adopted the Sobel test to check if the mediating effect is significant. As shown in Table 5, the Z value of the mediating effect of R&D investment passes the test at the 1% significance level, and the mediating effect of R&D investment accounts for 58.06%, once again indicating that IPR protection affects enterprise green innovation through the R&D investment channel.

    The empirical findings imply that the role of intellectual property protection for enterprise green innovation should be intensified from two aspects. One is to reinforce the rapid and coordinated protection of intellectual property. Local IPR management departments need to strengthen information communication, share relevant information of enterprises under their jurisdiction, collect information on IPR protection needs and case clues, and increase the efforts of administrative adjudication of patent infringement disputes among enterprises to provide rapid and coordinated protection of IPR. Another aspect is to strengthen the assistance of enterprises in IPR defence. Local IPR management departments should actively innovate the working mode, promote the extension of the work system of IPR protection and assistance at the grassroots level, and explore the development of special actions for IPR protection and assistance.

    5.2.   The mediating effect of financing constraints

    Having investigated the mechanism by which IPR protection affects enterprise green innovation, with R&D investment as the mediating variable, we further introduced financing constraints as a mediating variable to explore whether financing constraints also have a mediating effect in the process of IPR protection promoting the advancement of green innovation. The test results are presented in Table 6.

    Table 6.  The mediating effect of financing constraints.
    Plan A:Results of the mediating effect test
    (1) (2) (3)
    Explanatory variables Total effect Financing constraints The mediating effect of financing constraints
    Policy 0.0732** -0.00330*** 0.0549*
    (0.0320) (0.00117) (0.0309)
    FC -5.553***
    (0.677)
    Size 0.266*** 0.00608*** 0.300***
    (0.0232) (0.00109) (0.0243)
    Age 0.171 0.0290*** 0.332**
    (0.137) (0.00475) (0.136)
    Lev -0.0782 0.00779*** -0.0344
    (0.0649) (0.00289) (0.0624)
    ROA 0.0276 -0.00135 0.0211
    (0.124) (0.00389) (0.123)
    Board 0.0177 0.00426** 0.0422
    (0.0702) (0.00215) (0.0708)
    Indep -0.0693 0.00893 -0.0196
    (0.214) (0.00622) (0.214)
    SOE 0.169*** 0.000897 0.174***
    (0.0556) (0.00152) (0.0548)
    Growth -0.0203 0.00192*** -0.00970
    (0.0125) (0.000438) (0.0125)
    Constant -5.955*** 1.077*** 0.0244
    (0.648) (0.0243) (0.810)
    Controls YES YES YES
    Year YES YES YES
    Firm YES YES YES
    Observations 19,285 19,283 19,283
    Number of stkcd 2,321 2,321 2,321
    R-squared 0.154 0.846 0.167
    Plan B:Mediating effect test
    Mediating variable c a σa b σb Z Mediating effect weight
    Financing constraints 0.2152 0.0023 0.0011 -0.8134 0.1352 -2.003** -0.8769%
    Note that *, **, and *** indicate significant levels of 10%, 5%, and 1%, respectively; the standard errors of the regression coefficients are shown in parentheses.
     | Show Table
    DownLoad: CSV

    Based on the findings in Table 6, financing constraints have a negative impact in the process of IPR protection promoting enterprise green innovation. In other words, the smaller the financing constraints, the bigger the enterprise's green innovation capability; conversely, the greater the financing constraints, the more the green innovation capability will be inhibited. The regression findings in Column (2) show that the implementation of the intellectual property court pilot policy has a restraining impact on financing constraints. That is, financing constraints impede the development of green innovation, but the strengthening of IPR protection helps to mitigate the hindering impact of financing constraints on green innovation, thereby achieving the effect of promoting enterprise green innovation. One possible explanation is that strengthening IPR protection can increase the value of innovation achievements, thus allowing enterprises to obtain more financing [9]. The regression results presented in Column (3) demonstrate that even after adding financing constraints as a mediating variable with a negative effect, the DID coefficient of the policy remains significantly positive at 0.0549, indicating that the mitigating effect of IPR protection on financing constraints exists, and confirming Hypothesis 3.

    Similarly, to further verify whether financing constraints act as a mediating variable in the influence of IPR protection on green innovation in enterprises, the Sobel test was employed to check whether the mediating effect is significant. As shown in Table 6, the Z-value of the mediating effect of financing constraints has been found to be significant at the 5% level, and the weight of the mediating effect of financing constraints is -0.8769%, indicating once again that IPR protection promotes enterprise green innovation by mitigating the impact of financing constraints.

    According to the empirical conclusions, in order to alleviate financing constraints and enhance the effect of intellectual property protection on enterprise green technological innovation, policy guidance can be carried out in terms of three aspects: first is the acceleration of policy innovation that balances demand and supply. Policy innovation in the new era should adhere to systematic thinking. Second, in terms of the top-level design of the national innovation policy system, compared with supply-side support policies, it is more important to give play to the role of demand-side support policies. Third, it is important to enhance the pertinence of governmental policies for various types of enterprises and optimize the implementation methods and mechanisms of policies. Full consideration should be given to the differences in enterprise attributes, and targeted support should be implemented with precision.

    6.   Conclusions and policy implications

    This study conducted a natural experiment based on the pilot policy of the intellectual property courts in Beijing, Shanghai, and Guangzhou. By matching data from the CSMAR database of A-share listed companies and the National Intellectual Property Office from 2011 to 2019, the study selected the number of green invention patent applications filed by companies as a measure of their green innovation level. Utilizing a DID model and a mediating effect model, the study examined the effect of IPR protection on the development of enterprise green innovation and explored the impact mechanism.

    6.1.   Main conclusions

    First, strengthening IPR protection plays a significant role in promoting enterprise green innovation. After a series of robustness tests, such as the propensity score matching (PSM) and placebo tests, the conclusion still holds. IPR protection has a constructive impact on promoting industry transformation and achieving green development.

    Second, R&D investment plays a partial mediating role in promoting green innovation development through IPR protection. The establishment of the intellectual property pilot courts not only ensures enterprise IPR and better safeguards enterprise technical patent achievements, it also has a certain role in promoting the enterprise technology R&D, promoting the enterprise green innovation capabilities.

    Third, financing constraints have a negative mediating effect on the development of enterprise green innovation. Faced with investment uncertainty, investors usually use a company's patent technology achievements as a measure of the company's development potential. Higher quality patents often easily garner investment, and strengthening IPR protection can alleviate the financing constraints faced by enterprises and enhance their green innovation capabilities.

    Although this paper explores that local intellectual property protection has a positive effect on green technology innovation in enterprises and obtains channels of influence and some valuable conclusions, the heterogeneity of this impact can be further investigated.

    6.2.   Policy implications

    In order to better utilize the intellectual property protection system, create a favorable business environment, and promote green innovation and technological progress for enterprises, we propose the following policy implications.

    China should persist in strengthening and improving the pilot policies of the intellectual property court to promote the advancement of enterprise green innovation.

    The scope of the pilot program should be expanded, the judicial trial system for IPR protection should be reformed, and law enforcement capabilities should be improved. Besides, various policy measures should be adopted to stimulate the vitality of non-state-owned enterprises in green innovation.

    The government should give full play to its institutional and supportive role in intellectual property rights protection, assisting non-state-owned enterprises in creating a favorable intellectual property environment for green innovation.

    What's more, a long-term incentive mechanism should be established for green innovation. For the purpose of accelerating the implementation of the national innovation-driven development strategy, central and local governments have successively launched a series of innovation incentive policies and measures to promote enterprise innovation. All innovation incentive policies should focus on optimizing the green innovation environment and fully implement the concept of "lucid waters and lush mountains are invaluable assets" into policies.

    Use of AI tools declaration

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    Acknowledgments

    This research was funded by the National Social Science Fund Project "Research on Statistical Monitoring of Industrial Digitalization in China" (22BTJ053), the Key Project of the National Social Science Fund "Statistical Measurement and Optimization Path Research on China's Green Transformation of the Development Mode" (23ATJ007), the Youth Project of the Department of Education of Hunan Province (23B0831), and the Innovation and Entrepreneurship Projects for College Students (S202311528160).

    Conflict of interest

    The authors declare there is no conflict of interest.



    [1] X. Shao, H. Fan, G. Lu, J. Xu, An improved infrared dim and small target detection algorithm based on the contrast mechanism of human visual system, Infrared Phys. Technol., 55 (2012), 403–408. https://doi.org/10.1016/j.infrared.2012.06.001 doi: 10.1016/j.infrared.2012.06.001
    [2] M. Fan, S. Tian, K. Liu, J. Zhao, Y. Li, Infrared small target detection based on region proposal and CNN classifier, Signal, Image Video Process., 15 (2021), 1927–1936. https://doi.org/10.1007/s11760-021-01936-z doi: 10.1007/s11760-021-01936-z
    [3] Y. Zhang, L. Zheng, Y. Zhang, Small infrared target detection via a Mexican-Hat distribution, Appl. Sci., 9 (2019). https://doi.org/10.3390/app9245570 doi: 10.3390/app9245570
    [4] X. Cao, C. Rong, X. Bai, Infrared small target detection based on derivative dissimilarity measure, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 12 (2019), 3101–3116. https://doi.org/10.1109/JSTARS.2019.2920327 doi: 10.1109/JSTARS.2019.2920327
    [5] C. Chen, H. Li, Y. Wei, T. Xia, Y. Tang, A local contrast method for small infrared target detection, IEEE Trans. Geosci. Remote Sens., 52 (2014), 574–581. https://doi.org/10.1109/TGRS.2013.2242477 doi: 10.1109/TGRS.2013.2242477
    [6] J. Han, Y. Ma, B. Zhou, F. Fan, K. Liang, Y. Fang, A robust infrared small target detection algorithm based on human visual system, IEEE Geosci. Remote Sens. Lett., 11 (2014), 2168–2172. https://doi.org/10.1109/LGRS.2014.2323236 doi: 10.1109/LGRS.2014.2323236
    [7] Y. Qin, B. Li, Effective infrared small target detection utilizing a novel local contrast method, IEEE Geosci. Remote Sens. Lett., 13 (2016), 1890–1894. https://doi.org/10.1109/LGRS.2016.2616416 doi: 10.1109/LGRS.2016.2616416
    [8] L. Wu, Y. Ma, F. Fan, M. Wu, J. Huang, A double-neighborhood gradient method for infrared small target detection, IEEE Geosci. Remote Sens. Lett., 18 (2021), 1476–1480. https://doi.org/10.1109/LGRS.2020.3003267 doi: 10.1109/LGRS.2020.3003267
    [9] Y. Lu, L. Dong, T. Zhang, W. Xu, A robust detection algorithm for infrared maritime small and dim targets, Sensors, 20 (2020). https://doi.org/10.3390/s20041237 doi: 10.3390/s20041237
    [10] X. Wang, Z. Peng, D. Kong, Y. He, Infrared dim and small target detection based on stable multisubspace learning in heterogeneous scene, IEEE Trans. Geosci. Remote Sens., 55 (2017), 5481–5493. https://doi.org/10.1109/TGRS.2017.2709250 doi: 10.1109/TGRS.2017.2709250
    [11] L. Dong, B. Wang, M. Zhao, W. Xu, Robust infrared maritime target detection based on visual attention and spatiotemporal filtering, IEEE Trans. Geosci. Remote Sens., 55 (2017), 3037–3050. https://doi.org/10.1109/TGRS.2017.2660879 doi: 10.1109/TGRS.2017.2660879
    [12] L. Ding, X. Xu, Y. Cao, G. Zhai, F. Yang, L. Qian, Detection and tracking of infrared small target by jointly using SSD and pipeline filter, Digital Signal Process., 110 (2021). https://doi.org/10.1016/j.dsp.2020.102949 doi: 10.1016/j.dsp.2020.102949
    [13] Q. Guo, Z. Li, W. Song, W. Fu, Parallel computing based dynamic programming algorithm of track-before-detect, Symmetry, 11 (2019). https://doi.org/10.3390/sym11010029 doi: 10.3390/sym11010029
    [14] Q. Hu, J. Yang, P. Qin, S. Fong, Towards a context-free machine universal grammar (CF-MUG) in natural language processing, IEEE Access, 8 (2020). https://doi.org/10.1109/ACCESS.2020.3022674 doi: 10.1109/ACCESS.2020.3022674
    [15] A. Ilioudi, A. Dabiri, B. J. Wolf, B. De Schutter, Deep learning for object detection and segmentation in videos: toward an integration with domain knowledge, IEEE Access, 10 (2022), 34562–34576. https://doi.org/10.1109/ACCESS.2022.3162827 doi: 10.1109/ACCESS.2022.3162827
    [16] R. Girshick, J. Donahue, T. Darrell, J. Malik, Rich feature hierarchies for accurate object detection and semantic segmentation, in 2014 IEEE Conference on Computer Vision and Pattern Recognition, (2014), 580–587. https://doi.org/10.1109/CVPR.2014.81
    [17] R. Girshick, Fast R-CNN, in 2015 IEEE International Conference on Computer Vision (ICCV), (2015), 7–13. https://doi.org/10.1109/ICCV.2015.169
    [18] S. Ren, K. He, R. Girshick, J. Sun, Faster R-CNN: Towards real-time object detection with region proposal networks, IEEE Trans. Pattern Anal. Mach. Intell., 39 (2017), 1137–1149. https://doi.org/10.1109/TPAMI.2016.2577031 doi: 10.1109/TPAMI.2016.2577031
    [19] W. Liu, A. Dragomir, E. Dumitru, S. Christian, R. Scott, C. Fu, et al., SSD: Single shot multibox detector, in Lecture Notes in Computer Science, (2016), 21–37. https://doi.org/10.1007/978-3-319-46448-0_2
    [20] J. Redmon, A. Farhadi, YOLOV3: An incremental improvement, arXiv preprints, (2018), arXiv: 1804.02767. https://doi.org/10.48550/arXiv.1804.02767
    [21] T. Y. Lin, P. Goyal, R. Girshick, K. He, P. Dollár, Focal loss for dense object detection, in 2017 IEEE International Conference on Computer Vision (ICCV), (2017), 2999–3007. https://doi.org/10.1109/ICCV.2017.324
    [22] M. Ju, J. Luo, P. Zhang, M. He, H. Luo, A simple and efficient network for small target detection, IEEE Access, 7 (2019), 85771–85781. https://doi.org/10.1109/ACCESS.2019.2924960 doi: 10.1109/ACCESS.2019.2924960
    [23] H. Liang, J. Yang, M. Shao, FE-RetinaNet: small target detection with parallel multi-scale feature enhancement, Symmetry, 13 (2021). https://doi.org/10.3390/sym13060950 doi: 10.3390/sym13060950
    [24] Q. Hou, Z. Wang, F. Tan, Y. Zhao, H. Zheng, W. Zhang, RISTDnet: Robust infrared small target detection network, IEEE Geosci. Remote Sens. Lett., 19 (2021), 1–5. https://doi.org/10.1109/LGRS.2021.3050828 doi: 10.1109/LGRS.2021.3050828
    [25] Y. Li, S. Li, H. Du, L. Chen, D. Zhang, Y. Li. YOLO-ACN: Focusing on small target and occluded object detection, IEEE Access, 8 (2020), 227288–227303. https://doi.org/10.1109/ACCESS.2020.3046515 doi: 10.1109/ACCESS.2020.3046515
    [26] Z. Song, Y. Zhang, Y. Liu, K. Yang, M. Sun, MSFYOLO: Feature fusion-based detection for small objects, IEEE Lat. Am. Trans., 20 (2022), 823–830. https://doi.org/10.1109/TLA.2022.9693567 doi: 10.1109/TLA.2022.9693567
    [27] K. Wang, S. Li, S. Niu, K. Zhang, Detection of infrared small targets using feature fusion convolutional network, IEEE Access, 7 (2019), 146081–146092. https://doi.org/10.1109/ACCESS.2019.2944661 doi: 10.1109/ACCESS.2019.2944661
    [28] M. Ju, J. Luo, G. Liu, H. Luo, ISTDet: An efficient end-to-end neural network for infrared small target detection, Infrared Phys. Technol., 114 (2021). https://doi.org/10.1016/j.infrared.2021.103659 doi: 10.1016/j.infrared.2021.103659
    [29] S. Yao, Q. Zhu, T. Zhang, W. Cui, P. Yan, Infrared image small-target detection based on improved FCOS and spatio-temporal features, Electronics, 11 (2022). https://doi.org/10.3390/electronics11060933 doi: 10.3390/electronics11060933
    [30] X. Tong, B. Sun, J. Wei, Z. Zuo, S. Su, EAAU-Net: Enhanced asymmetric attention U-Net for infrared small target detection, Remote Sens., 13 (2021). https://doi.org/10.3390/rs13163200 doi: 10.3390/rs13163200
    [31] Q. Xu, H. Huang, C. Zhou, X. Zhang, Research on real-time infrared image fault detection of substation high-voltage lead connectors based on improved YOLOv3 network, Electronics, 10 (2021). https://doi.org/10.3390/electronics10050544 doi: 10.3390/electronics10050544
    [32] C. Deng, M. Wang, L. Liu, Y. Liu, Y. Jiang, Extended feature pyramid network for small object detection, IEEE Trans. Multimedia, 24 (2021), 1968–1979. https://doi.org/10.1109/TMM.2021.3074273 doi: 10.1109/TMM.2021.3074273
    [33] G. Huang, Z. Liu, L. Van Der Maaten, K. Q. Weinberger, Densely connected convolutional networks, in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (2017), 2261–2269. https://doi.org/10.1109/CVPR.2017.243
    [34] Y. Dai, Y. Wu, F. Zhou, K. Barnard, Asymmetric contextual modulation for infrared small target detection, in 2021 IEEE Winter Conference on Applications of Computer Vision (WACV), (2021), 949–958. https://doi.org/10.1109/WACV48630.2021.00099
    [35] Q. Wang, B. Wu, P. Zhu, P. Li, W. Zuo, Q. Hu, ECA-Net: Efficient channel attention for deep convolutional neural networks, in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), (2020), 11531–11539. https://doi.org/10.1109/CVPR42600.2020.01155
    [36] M. Pastorino, G. Moser, S. B. Serpico, J. Zerubia, Fully convolutional and feedforward networks for the semantic segmentation of remotely sensed images, in 2022 IEEE International Conference on Image Processing (ICIP), (2022), 1876–1880. https://doi.org/10.1109/ICIP46576.2022.9897336
    [37] M. Lin, Q. Chen, S. Yan, Network in network, arXiv preprints, (2013), arXiv: 1312.4400. https://doi.org/10.48550/arXiv.1312.4400
    [38] B. Hui, Z. Song, H. Fan, P. Zhong, W. Hu, X. Zhang, et al., A dataset for infrared image dim-small aircraft target detection and tracking under ground/air background, Science Data Bank, (2019). https://doi.org/10.11922/sciencedb
    [39] X. Zhou, D. Wang, P. Krähenbähl, Objects as points, arXiv preprint, (2019), arXiv: 1904.07850. http://arXiv.org/abs/1904.07850

  • This article has been cited by:

    1. Xiangsong Ye, Longju Zhou, Zhiyong Wang, Spatial convergence characteristics of low carbon economy and economic growth quality: based on Guangdong urban data, 2024, 9, 2473-6988, 11660, 10.3934/math.2024572
    2. Yue Liu, Yixian Wen, Yulan Xiao, Lichang Zhang, Shan Huang, Identification of the enterprise financialization motivation on crowding out R&D innovation: evidence from listed companies in China, 2024, 9, 2473-6988, 5951, 10.3934/math.2024291
    3. Zhongzhe Ouyang, Ke Liu, Min Lu, Bias correction based on AR model in spurious regression, 2024, 9, 2473-6988, 8439, 10.3934/math.2024410
    4. Jiajia He, Xiuping Zou, Tinghui Li, House price, gender spatial allocation, and the change of marriage matching, 2024, 9, 2473-6988, 8079, 10.3934/math.2024393
    5. Zimei Huang, Zhenghui Li, The impact of digital economy on the financial risk ripple effect: evidence from China, 2024, 9, 2473-6988, 8920, 10.3934/math.2024435
    6. Yue Liu, Shan Huang, Zirui Wang, Lichang Zhang, Research on the moderate range of enterprise financialization and R&D innovation from a nonlinear perspective, 2024, 32, 2688-1594, 1923, 10.3934/era.2024088
    7. Qiong Liu, Rujun Xu, Wenli Tao, Impact of auditing information technology modernization on corporate technological innovation investment: The roles of risk perception and internal control, 2024, 67, 15446123, 105807, 10.1016/j.frl.2024.105807
    8. Qingzhen Deng, Liguang Zhang, Shasha Wang, Yunxiang Liao, Jiaqi Zeng, State‐owned equity participation and corporate green innovation: evidence from Chinese private enterprises, 2024, 0818-9935, 10.1111/apel.12435
    9. Zhenghui Li, Hanzi Chen, Siting Lu, Pierre Failler, How does digital payment affect international trade? Research based on the social network analysis method, 2024, 32, 2688-1594, 1406, 10.3934/era.2024065
    10. Changfei Nie, Ziyan Qi, Yuan Feng, Intellectual property judicial protection, innovation factor input, and corporate innovation: Evidence from the establishment of intellectual property courts in China, 2024, 45, 0143-6570, 4227, 10.1002/mde.4255
    11. Qifeng Han, Chunhui Li, Yulu Jin, The impact of intellectual property protection on the development of artificial intelligence in enterprises, 2025, 10575219, 104179, 10.1016/j.irfa.2025.104179
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索
Mathematical Biosciences and Engineering
4.4

Metrics

Article views(2643) PDF downloads(127) Cited by(2)

Article has an altmetric score of 1
Article has an altmetric score of 1
Preview PDF
Download XML
Export Citation
Article outline
Abstract
   Introduction
   Research scheme
   Indicator selection and data sources
   Econometric tests of the impact of IPR protection on enterprise green innovation
   The impact mechanism of the IPR protection on enterprise green innovation
   Conclusions and policy implications
Use of AI tools declaration
Acknowledgments
Conflict of interest
References

Figures and Tables

Figures(14)  /  Tables(4)

Other Articles By Authors

Related pages

Tools

Copyright © AIMS Press

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog