Research article Topical Sections

Major technological differences between an industrial-type and five artisan-type Greek PDO Galotyri market cheeses as revealed by great variations in their lactic acid microbiota

  • Galotyri is a Greek PDO acid-curd, traditionally ripened, soft spread cheese produced from boiled (>80–90 ℃), naturally acidified ewes’ or ewes’/goats’ milk in the regions of Epirus and Thessaly. In this survey study, the microbiology of an industrial Galotyri cheese variety (11 batches) was compared with five artisan-type Galotyri cheese varieties (1–2 batches each), with emphasis on the species of lactic acid bacteria (LAB) prevailing in the final product. All market cheese samples distributed either in packaged or bulk form, were purchased from retail outlets at Ioannina, Epirus during two sampling periods, 2003–2006 and 2014–2019. A total of 366 (189 industrial and 177 artisan) cheese LAB isolates were identified by key biochemical criteria. Strains of Streptococcus thermophilus and/or Lactobacillus delbrueckii subsp. bulgaricus were isolated nearly as a pure symbiotic starter culture (97.4%) from all industrial batches, labeled as a fresh cheese combining the nutritional properties of yogurt. Conversely, the artisan-type cheeses were ripened and more diversified in terms of LAB ecology. They contained strains of Lactococcus lactis (19.8%), Lb. plantarum (16.9%), S. thermophilus (14.7%), Leuconostoc mesenteroides group (11.3%), Pediococcus inopinatus (7.3%), Enterococcus faecalis (8.5%), E. faecium (6.2%), E. durans (5.1%), Lb. rhamnosus (3.4%), Lb. gasseri (3.4%), Lb. bulgaricus (2.8%) and Lb. acidipiscis (0.6%); the prevalence of each species was cheese product batch-dependent. In conclusion: (i) the industrial Galotyri cheese technology results in a fresh yogurt-like cheese of limited (thermophilic) LAB diversity; (ii) the artisan-type Galotyri cheese technologies may vary significantly, but share a decisive ripening step that favors the predominance of mesophilic LAB species in the final market products; (iii) a regulatory distinction between fresh and ripened varieties of commercial PDO-certified Galotyri cheeses is required.

    Citation: John Samelis, Athanasia Kakouri. Major technological differences between an industrial-type and five artisan-type Greek PDO Galotyri market cheeses as revealed by great variations in their lactic acid microbiota[J]. AIMS Agriculture and Food, 2019, 4(3): 685-710. doi: 10.3934/agrfood.2019.3.685

    Related Papers:

    [1] Raimund Bürger, Paola Goatin, Daniel Inzunza, Luis Miguel Villada . A non-local pedestrian flow model accounting for anisotropic interactions and domain boundaries. Mathematical Biosciences and Engineering, 2020, 17(5): 5883-5906. doi: 10.3934/mbe.2020314
    [2] Wenshun Sheng, Jiahui Shen, Qiming Huang, Zhixuan Liu, Zihao Ding . Multi-objective pedestrian tracking method based on YOLOv8 and improved DeepSORT. Mathematical Biosciences and Engineering, 2024, 21(2): 1791-1805. doi: 10.3934/mbe.2024077
    [3] Enrico Bertino, Régis Duvigneau, Paola Goatin . Uncertainty quantification in a macroscopic traffic flow model calibrated on GPS data. Mathematical Biosciences and Engineering, 2020, 17(2): 1511-1533. doi: 10.3934/mbe.2020078
    [4] Wen Li, Xuekun Yang, Guowu Yuan, Dan Xu . ABCNet: A comprehensive highway visibility prediction model based on attention, Bi-LSTM and CNN. Mathematical Biosciences and Engineering, 2024, 21(3): 4397-4420. doi: 10.3934/mbe.2024194
    [5] Alessandro Corbetta, Adrian Muntean, Kiamars Vafayi . Parameter estimation of social forces in pedestrian dynamics models via a probabilistic method. Mathematical Biosciences and Engineering, 2015, 12(2): 337-356. doi: 10.3934/mbe.2015.12.337
    [6] Veronika Schleper . A hybrid model for traffic flow and crowd dynamics with random individual properties. Mathematical Biosciences and Engineering, 2015, 12(2): 393-413. doi: 10.3934/mbe.2015.12.393
    [7] Fengjie Liu, Monan Wang, Yuzheng Ma . Multiscale modeling of skeletal muscle to explore its passive mechanical properties and experiments verification. Mathematical Biosciences and Engineering, 2022, 19(2): 1251-1279. doi: 10.3934/mbe.2022058
    [8] Paola Goatin, Matthias Mimault . A mixed system modeling two-directional pedestrian flows. Mathematical Biosciences and Engineering, 2015, 12(2): 375-392. doi: 10.3934/mbe.2015.12.375
    [9] H. T. Banks, R. A. Everett, Neha Murad, R. D. White, J. E. Banks, Bodil N. Cass, Jay A. Rosenheim . Optimal design for dynamical modeling of pest populations. Mathematical Biosciences and Engineering, 2018, 15(4): 993-1010. doi: 10.3934/mbe.2018044
    [10] Simone Göttlich, Stephan Knapp . Modeling random traffic accidents by conservation laws. Mathematical Biosciences and Engineering, 2020, 17(2): 1677-1701. doi: 10.3934/mbe.2020088
  • Galotyri is a Greek PDO acid-curd, traditionally ripened, soft spread cheese produced from boiled (>80–90 ℃), naturally acidified ewes’ or ewes’/goats’ milk in the regions of Epirus and Thessaly. In this survey study, the microbiology of an industrial Galotyri cheese variety (11 batches) was compared with five artisan-type Galotyri cheese varieties (1–2 batches each), with emphasis on the species of lactic acid bacteria (LAB) prevailing in the final product. All market cheese samples distributed either in packaged or bulk form, were purchased from retail outlets at Ioannina, Epirus during two sampling periods, 2003–2006 and 2014–2019. A total of 366 (189 industrial and 177 artisan) cheese LAB isolates were identified by key biochemical criteria. Strains of Streptococcus thermophilus and/or Lactobacillus delbrueckii subsp. bulgaricus were isolated nearly as a pure symbiotic starter culture (97.4%) from all industrial batches, labeled as a fresh cheese combining the nutritional properties of yogurt. Conversely, the artisan-type cheeses were ripened and more diversified in terms of LAB ecology. They contained strains of Lactococcus lactis (19.8%), Lb. plantarum (16.9%), S. thermophilus (14.7%), Leuconostoc mesenteroides group (11.3%), Pediococcus inopinatus (7.3%), Enterococcus faecalis (8.5%), E. faecium (6.2%), E. durans (5.1%), Lb. rhamnosus (3.4%), Lb. gasseri (3.4%), Lb. bulgaricus (2.8%) and Lb. acidipiscis (0.6%); the prevalence of each species was cheese product batch-dependent. In conclusion: (i) the industrial Galotyri cheese technology results in a fresh yogurt-like cheese of limited (thermophilic) LAB diversity; (ii) the artisan-type Galotyri cheese technologies may vary significantly, but share a decisive ripening step that favors the predominance of mesophilic LAB species in the final market products; (iii) a regulatory distinction between fresh and ripened varieties of commercial PDO-certified Galotyri cheeses is required.


    In an information era, owing to hardware and software advances, the development of technologies in engineering is tented to high computational ability. Intelligent computing and machine learning provide a good strategy to solve the complicated problems and a fast computational ability to reduce the complexity of the procedure. In the computer vision field, it usually operates in the frequency domain to extract the essential information in order to model the events. However, the procedure operated in the frequency domain is a kind of complex computation. Involving the intelligence into the modeling technique to solve the confronted problems of computer vision is the current trend. Intelligent computing is the integration of many technologies and decisions of knowledge for the computing environment. Computational photography scheme, image processing, requests the intelligent computing to deal with the object model and feature extraction. Motion blur is usually presented in many practical scenarios, such as hand-held cameras especially mobile material. The motion-blurred picture is caused by the relative motion between the camera and an imaged scene during exposure. Generally, the relative motion can be divided into two kinds: camera shake and object motion. Sensor movement during exposure leads to unwanted blur in the acquired image. Assuming a scene and ignoring the effects of defocus and lens abnormality, each point in the blurred image can be modeled as the convolution of the unblurred image by a global point spread function (PSF). Image deblurring aims to achieve a deconvolution process to recover the clear image from the acquired blurry image. Based on intelligent computing strategy, image deblurring could be reduced the computing complexity and improved the computational time.

    Artificial neural networks (NNs) have been used extensively in image processing [1,2]. Schuler et al. [3] proposed a learning-based neural network (NN) to estimate the features and then estimate blur kernel for deblurring by using the deconvolution. Additionally, a common assumption in motion deblurring methods is that the motion PSF is spatially invariant. This implies that all pixels are convolved with the same motion blur kernel. The problem of blur kernel estimation and more generally blind deconvolution is a long-existing problem in computer vision. Restoration of blurry images is highly dependent on estimation of motion blur kernel after implementing the appropriate image restoration method. Many well-known PSF algorithms to estimate blur kernel have been proposed [4,5,6,7,8]. When the PSF is known, or can be estimated, a deconvolution algorithm, such as Richardson-Lucy [9,10,11,12], can be used to deblur the image. Motion blur is a type of the relative motion of the camera and shooting scene during exposure. Mathematically, the corresponding motion blur information is usually modeled as a linear image degradation process.

    B=IK+N, (1.1)

    where denotes the convolution operator, B, I, K, and N denote the blurred image, true sharp image, unknown blur kernel, and noise term, respectively. Blind image deconvolution is an inherently ill-posed problem since the blurred image B does not provide enough information for determining both I and K. Therefore, how to estimate blur kernel K from blurred image B is an important issue in motion-blurred image restoration. Many of studies have been proposed. Some studies are briefly described in the following.

    Blurred kernel information is usually hidden in the regions with edges, if the edge of an image suffered serious damage, it will result in inaccurate blur kernel estimation. Tai et al. [10] proposed a modified Richardson-Lucy (RL) method to incorporate space-invariant blur model under a projective motion path for image scenes. Yang et al. [11] proposed blur kernel estimation and non-blind image deconvolution to deblur image by using bilateral filter and Gradient attenuation Richardson-Lucy deconvolution algorithm. Dobeš et al. [13] and Goldstein and Fattal [5] proposed the kernel estimation in frequency domain. The blur kernel is then recovered using a phase retrieval algorithm with improved convergence and disambiguation capabilities. Deblurring approaches, based on spectral properties and edge information of an image have presented by [5,14,15] to retrieval the blur kernel information. In addition, one of many deblurring techniques is to incorporate image priors to impose on the deblurred results. Deshpande and Patnaik [16] proposed an image motion deblurring algorithm based on dual Fourier spectrum combined with bit plane slicing algorithm and Radon transform (RT) for accurate estimation of PSF parameters (blur length and blur angle). Shao et al. [17] used non-stationary Gaussian prior to estimate the salient edges of image as the cues to blur kernel estimation. He et al. [18] proposed motion blurring which used different priors for the local region and the motion blur kernel to formulate a minimization energy function that alternates between blur kernel estimation and deblurring image restoration. Jia [19] relied on color mixtures to estimate the motion blur kernel of moving objects given their boundary alpha values. Levin et al. [20] used a maximum a posteriori estimation (MAP) to estimate blur kernel and achieve the deblurring results. Except MAP methods [21,22], many methods are being developed [13,23,24,25,26].

    In this paper, we propose a motion deblurring method based on fast PSF (FPSF) to achieve image restoration. The advantages of this system can speed up the running time and find an optimal blur kernel, as well as obtain a good image quality for deblurring. In addition, in order to verify the reliability of our proposed system, the experimental data include the real motion-blurred images and artificial blurred images.

    The rest of this paper is organized as follows. Section 2 describes the related techniques. Section 3 describes the proposed method which includes blur kernel clustering, blur kernel integration, and the optimal blue kernel searching. Section 4 presents the experimental results. Finally, a conclusion is given in Section 5.

    In this section, we briefly describe the techniques related to our proposed approach.

    A particular property of natural image scenes can be illustrated by the following power-law relationship [27,28].

    ˆI(ω)2ωβ, (2.1)

    where ω is the coordinates in frequency domain, ˆI denotes Fourier transform of a natural image (I), according to the literature, β2.

    As the researches indicated, the blur information is hidden in the power-law of the neighborhoods of edges, therefore, filtering an image forming Eq. (2.1) can be acquired this information and is expressed as

    |^Id(ω)|2=|ˆI(ω)|2|ˆd(ω)|2π2ω2ω2=c, (2.2)

    for constant c, d is a first-order Laplacian filter. Thus, a blurry image B=Ik, this filtering process can be used to estimate the following blur-kernel power spectrum |ˆk(ω)|2.

    |^(Bd)(ω)|2=|ˆI(ω)|2|ˆd(ω)|2|ˆk(ω)|2c|ˆk(ω)|2. (2.3)

    The power spectrum of any signal F can relay to its autocorrelation according to the Wiener-Khinchin theorem [29].

    ˆRF(ω)=|ˆF(ω)|2, (2.4)

    where the autocorrelation is defined by RF(x)=(ˉFF)(x). The blur approximation in Eq. (2.3) can be identified by real-space parts for the spectrum and can be expressed as

    RBd(x)cRk(x). (2.5)

    Evidently, the power spectrum of a natural image varies by multiplicative factors along the different directions, that is,

    |ˆI(ω)|2cθ(ω)ω2, (2.6)

    where θ(ω)=arctan(ωx,ωy) is the angle of the vector ω.

    Blur kernel information, like cθ(ω) and the kernel phase, can be recovered by means of the Fourier slice theorem and Wiener-Khinchin theorem, given autocorrelation functions computed from the input blurry image B(x). From Eq. (2.6) known, only single parameter cθ is unknown. Based on Fourier theorem and Wiener-Khinchin theorem, Eq.(2.3) can be rewritten in real-space and expressed as Eq.(2.7) [5].

    fθ(x)cθRPθ(k)(x),θ[π,π], (2.7)

    where Pθ is a projection of a 2D signal into 1D by integrating it along the direction orthogonal to θ. By repeating this procedure for all the θ, an approximation for the 2D blur-kernel power spectrum function |ˆk(ω)|2 can be obtained. In [5], this procedure repeated three times. In addition, based an iterative phase-retrieval algorithm, this approximation can recover the blur kernel k.

    The phase recovery also called phase retrieval. As described above, recovering the kernel k, given its power spectrum |ˆk|2 requires estimating the phase component of ˆk(ω). However, this procedure only obtains the spectrum information, the phase information is still unknown because it iteratively switches between Fourier and real-space domains. In addition, the input |ˆk|2 and the spatial constraints may not guarantee a unique solution. Moreover, as [30] discussed, it may converge to local minima formation, therefore, the phase retrieval procedure is repeated multiple times under starting randomly phase component to further estimate the blur kernel.

    According to GS algorithm [31], this algorithm is a common method for phase retrieval. It is based on iterative Fourier transform and inverse Fourier transform between the object domain and the Fourier domain. A hybrid input-output method is used to estimate the blur kernel in iterative phase retrieval procedure under the appropriate frequency/spatial domain constraints [5,31]. Therefore, based on iterative phase retrieval algorithm, the blur kernel can be recovered. Thus, the blurry image can be deblurred through a deconvolution. The procedures of phase retrieval algorithm are briefly described as follows. Its pseudo code is shown in Algorithm 1 (see [5] for details).

    Step 1) Given the initial phase ϕ(ω) within [π,π] randomly generated.

    Step 2) Transfrom the real-space domain g using inverse Fourier transform.

    Step 3) Transform g into ˆg using Fourier domain constraints.

    Step 4) Transform to real-space g2 using the phase information of ˆg.

    Step 5) Compute the constraint R(x) using the space domain constraints.

    Step 6) Obtain a hybrid input-output constraint Ω which is the union of R(x) and s.

    Step 7) Repeat Steps 3 - 6 for m times, and output the kn based on the constraint Ω which is the union of g2(x) and s.

    Algorithm 1 Iterative phase retrieval
    1: Input: kernel magnitude spectrum, p(ω)=∣ˆk(ω) and kernel size =s
    2: for n=0 to Nguesses do
    3:   //initiate the phase ϕ(ω) randomly;
    4:   Sample ϕ(ω) uniformly from [π,π]
    5:   //transform to real space using inverse Fourier transform
    6:   g=F1(peiϕ)
    7: for m=1 to Ninner do
    8:     //appply fourier domain constraints;
    9:     g2=F1((αp+(1α)ˆg)expiϕ(ˆg))
    10:     //apply space domain constraints
    11:     R(x)=2g2(x)g(x)
    12:     β=β0+(1β0)(1exp(m/7)3)
    13:     Ω={x:R(x)<0}{x:x[0,s]×[0,s]}
    14:   g(x)={βg(x)+(12β)g2(x),ifxΩ,g2(x),ifxΩ;
    15:   end for
    16:   Ω={x:g2(x)<0}{x:x[0,s]×[0,s]}
    17:   kn(x)={0,ifxΩ,g2(x),ifxΩ
    18:   end for
    19:   Output: kn

    However, using the phase retrieval method, it cannot guarantee to obtain the same kernel in each of iterations and cannot decide which one is the best blur kernel, as shown in Fig. 1. Figure 1 shows thirty blur kernels which include the symmetric or failure blur kernels after performing thirty iterations in phase retrieval. Assuming that repeating phase retrieval for n times, it can find n blur kernels. Afterwards, an optimal kernel can be obtained from these n kernels. [5] takes the final kernel to deconvolve the blurry image. In our experiments, we utilize a normalized sparsity measure (NSM) [24] to decide the optimal blur kernel from n blur kernels and to estimate the quality of blur kernel.

    Figure 1.  Results of iterative phase retrieval for 30 iteration times placed at inverse s-shaped arrangement. Each kernel is of size 21 × 21.

    As known above, it can obtain n blur kernels after iterating n times. Each of NSM values for deconvolution by using the corresponding kernel can be calculated. Figure 1 shows an example of blur kernel for thirty iteration times. From Fig. 1, it is obvious that the symmetric relationship exists among blur kernels and the estimated blur kernel for each iteration is also different. Hence, the measure of blur kernel quality will test the symmetry of blur kernel and will make a score. In order to estimate the symmetric characteristic of blur kernel, it has to calculate the NSM score twice. For example, if there are thirty kernels, it will calculate the NSM for sixty times. After computing NSM value, the smaller the NSM score, the better the reconstructed image, as shown in Fig. 2. According to our experimental results, a kernel with the minimum NSM value is a good kernel more confidently.

    Figure 2.  NSM value (an example). (a)Blurry image. (b)Reconstructed image.

    In our system, we modify the number of computing NSM times and thus speed up an optimal blur kernel acquirement from n blur kernels.

    Natural image signals are highly structural information, such as pixels exhibiting strong dependencies and containing important information about the structure of the objects in the visual scene. In order to estimate the structural performance of the reconstructed image after deconvolution, we adopt the structural-similarity-based image quality measure (SSIM) [32] instead of the mean squared error (MSE). The SSIM mainly compute the structural similarity between the reference and the distorted signals. However, one usually requires the overall image quality measure, a mean SSIM (MSSIM) derived from SSIM is used to achieve this measure, which can exhibit much better consistency with the qualitative visual appearance. In our experiments, we adopt the MSSIM to estimate the quality of the reconstructed image. The MSSIM index is briefly defined as

    MSSIM(X,Y)=1MMj=1SSIM(xj,yj), (2.8)

    where X and Y are the reference and the distorted images, respectively; xj and yj are the image contents at the jth local window; and M is the number of local windows of the image. The higher the MSSIM, the closer both images.

    In this paper, we propose an image deblurring using fast point spread function (FPSF) method to efficiently and quickly search the optimal blur kernel. Because the method by [5] is time-consuming, in our approach, we will further improve the computational time and speed up decide the optimal blur kernel to deblur the blurred image. Figure 3 illustrates the flowchart of the proposed system. The details of the procedures are described in the following.

    Figure 3.  The diagram of system flowchart.

    In order to improve the blur kernel estimation which is seriously affected by high-frequency components [5], we use Gaussian filter [33] to reduce this influence and improve the blur kernel estimation for a blurred image. Here, we use Gaussian filter with size 5×5, σ=0.6, and a mean of zero, it is proportional to the size of the neighborhood on which the filter operates. Pixels more distant from the center of the operator have smaller influence. After filtering the blurry image, it can improve the power spectrum of 2D blur kernels, afterwards, the optimal blur kernel can be acquired by our proposed method.

    After doing blur kernel estimation and iterative phase retrieval described by the previous section, it can obtain the corresponding kernels in each of iterative times in phase retrieval. However, from Figure 1, it is clearly obvious that each of these kernels is different. In other words, it does not guarantee that these kernels can yield a good deblurring result. Based on this reason, we propose a FPSF method to estimate an optimal blur kernel. Using this estimated optimal blur kernel, it can deblur the blurry image by deconvolution. Figure 4 illustrates the flowchart of the FPSF.

    Figure 4.  The block diagram of the FPSF.

    This method consists of blur kernel clustering and blur kernel integration. The blur kernel clustering based on MSSIM is to classify these blur kernels obtained by phase retrieval. After performing the clustering, we use the integration to find the optimal kernel from the clusters. For clustering, we use the MSSIM to estimate the similarity of all candidate kernels, kernel which has a high MSSIM will be clustered into the same cluster. The procedures of the clustering method are described as follows.

    First of all, assuming that there are kn blur kernels.

    Step 1) Select the first kernel in these kernels as an initial base.

    Step 2) Calculate the MSSIM values for this base and the rest of candidate kernels.

    Step 3) If MSSIM value is more than or equal to a threshold, then this candidate kernel is classified into the corresponding cluster.

    Step 4) If MSSIM value is less than a threshold, then a new base is yielded and it will serve as a new cluster.

    Step 5) Repeat Steps 2 - 4, until all candidate kernels are clustered.

    Based on MSSIM characteristic, the higher the MSSIM value, more similar both kernels. According to our experiments, this threshold is set 0.9. An example of kernel clustering algorithm is illustrated as follows. Assuming there are nine blur kernels, they will be clustered.

    First, kernel k1 is chosen as a base, next, the MSSIM value is computed. The first cycle can obtain the first cluster called k1 which includes k2, k6, and k8 kernels. Then, kernel k3 is a new cluster. After computing the clustering procedures, all candidate kernels can be classified into the corresponding clusters.

    Based on above clustering process, assume there are n blur kernels, they will be classified into Cm kernel clusters where mn. As described above, a good kernel can be obtained after computing the NSM for sixty times, but, it is very time-consuming. Therefore, in order to reduce the number of computing NSM times, we propose a blur kernel integration technique to gain this performance. This kernel integration technique consists of mean blur kernel calculation and refining described as follows.

    Mean blur kernel: We calculate an average blur kernel corresponding to each of kernel clusters with the relative coordinates. The average blur kernel for each of kernel clusters is defined as

    Km,avg(x,y)=1ggi=1Cm(i)(x,y), (3.1)

    where g denotes the number of kernels corresponding to kernel cluster, m denotes the number of kernel clusters, Cm()(x,y) denotes a kernel value corresponding to the x and y coordinates at the mth kernel cluster.

    Refining: Owing to the difference between kernels in cluster, the average kernel may involve noise (as shown in Figure 5 number 1's result). As described above, the noise will affect the deblurring process, hence, in order to reduce the influence of the noise and keep the important information of the kernel, we further refine the mean blur kernel to gain performance. We make a weight matrix to achieve the refining. First, we define a weight matrix of a kernel cluster which represents the number of nonzero kernel values corresponding to the coordinates in the same kernel cluster, the weight matrix is defined as

    Figure 5.  Example of kernel integration and its refining in kernel cluster. Number 1: computing an average kernel. Number 2: refining.
    wm(x,y)={wm(x,y)+1,ifCm(i)(x,y)>0,wm(x,y),otherwise,  (3.2)

    where the initial wm(x,y) sets to zero for m clusters. An average weight value is computed by

    aveWm=xywm(x,y)Sm, (3.3)

    where Sm is the number of nonzero kernel values corresponding to the relative cluster. The refining mean blur kernel is expressed as

    RKm,avg(x,y)={Km,avg(x,y),ifwm(x,y)aveWm,Km,avg(x,y)=0,otherwise. (3.4)

    Figure 5 illustrates an example of kernel integration processing in a kernel cluster.

    Based on the blur kernel integration, we can obtain a refining mean blur kernel corresponding to each of kernel clusters. Afterwards, using the measure of blur kernel quality again described by subsection 2.3 for m refining mean blur kernels, we obtain the best mean blur kernel and its corresponding cluster. Because we have recorded the symmetry of the cluster in searching cluster process, thus, we can greatly reduce the computational time and quickly find an optimal blur kernel. The procedures of searching an optimal blur kernel (OBK) are described as follows.

    Step 1) Give a set of blur kernels which belong to the best kernel cluster, and the corresponding symmetric property of the phase.

    Step 2) Deconvolve all the blurred images using these kernels.

    Step 3) Compute the relative NSM values for all deconvolution results.

    Step 4) Find the minimum NSM value from these NSM values and its corresponding kernel, this kernel serves as an optimal kernel.

    To verify the performance of our proposed method, the experimental results are compared with Goldstein and Fattal [5] and Krishnan et al. [24] to perform the visual quality of the reconstruction image and computational time. Based on the above procedures, we can obtain the optimal kernel for the corresponding blurry image, then each of three components (red, green, blue) for the color image is individually used this optimal kernel to work the non-blind deconvolution method based on [34]; finally, all blurry color images can be reconstructed.

    All experiments are implemented in Microsoft Visual Studio 2010 C, an Intel © i5 dual Core 3.2 GHz, and 4 GB RAM computer with window 7 64 bits platform.

    First, we need to make the experimental data by using predefined blur kernels. How to decide the number of predefined blur kernels is a trade-off between objectivity and variety of the experimental data. In order to perform these attributes, in our experiments, we make ten blur kernels with size 21×21 shown in Figure 6 to simulate ten kinds of blurred images. A database including ten images with size 768×1024, 682×1024, 1024×682, and 720×960 pixels separated 4, 3, 2, and 1 is given. Then we blurred each of these images using ten blur kernels, it can obtain 100 motion-blurred images for test. In order to avoid noise disturbance to affect kernel estimation, all of experimental data are firstly filtered by Gaussian filter. After filtering, we can obtain more convergent blur kernel by means of our proposed method. Here, we use Gaussian filter with the standard deviation σ=0.6 and run thirty iterations for phase retrieval to demonstrate our experiments.

    Figure 6.  Ten blur kernels used to simulate motion-blurred image. The inverse S-order denotes these kernels from k1 to k10.

    For performance measure, we use the peak signal-to-noise ratio (PSNR) and MSSIM to estimate the performance of the system. The PSNR and mean square error (MSE) are defined as

    PSNR=10logS2MaxMSE,MSE=1h×whiwj|I1(i,j)I2(i,j)|2, (4.1)

    where I1(i,j) and I2(i,j) denote the reconstructed image and the original one corresponding to the coordinates (i,j). h and w denote the height and width of the image, respectively. For a gray-level image, SMax is 255 gray value.

    Because our proposed method improved Goldstein and Fattal method, the power spectrum of blur kernel for each iteration in phase retrieval may be not equal. Hence, in order to obtain the stable performance in our experiments, each blurry image is doing blur kernel estimation and its convolution for ten times. In addition, the size of blur kernel will seriously affect the execution time and reconstructed image quality, therefore, we test the different kernel sizes to demonstrate the execution speed. First of all, we take an average execution time in which the blurry image is doing blur kernel estimation with the different kernel sizes for ten iterations. Figure 7 shows the average execution time for the different blur kernel sizes in our experimental data. From Figure 7, it is clear that the execution time increases with increasing kernel size, and our proposed method is more upgraded than those of methods. Hence, considering the computational time and the visual quality of the reconstructed images, we adopt the blur kernel of size 21 × 21.

    Figure 7.  The size of blur kernel and its computational time in seconds compared with the proposed method, Goldstein and Fattal's method, and Krishnan et al.'s method.

    Table 1 illustrates the PSNR value and MSSIM value. Image deblurring results of a part of tests are shown in Figures 8-9. Besides man-made test data, we also use the real motion-blurred images to demonstrate our proposed method, as shown in Figures 10-11.

    Table 1.  The PSNR(dB)/MSSIM values with the different deblur kernels compared with the proposed method, Goldstein and Fattal's [5] method, and Krishnan et al.'s [24] method.
    KernelMethod
    Proposed [5] [24]
    k122.13/0.9422.15/0.8120.47/0.70
    k223/0.9623.24/0.8520.36/0.73
    k323.95/0.9624.36/0.5325.94/0.85
    k422.6/0.9522.47/0.8322.04/0.79
    k521.69/0.9321.48/0.7920.16/0.65
    k622.42/0.9422.49/0.8222.8/0.83
    k719.43/0.8819.33/0.6818.77/0.57
    k824.27/0.9524.28/0.9521.16/0.88
    k922.13/0.9422.14/0.8021.63/0.78
    k1023.4/0.9523.54/0.8220.23/0.66
    Ave.22.5/0.9422.55/0.8221.36/0.75

     | Show Table
    DownLoad: CSV
    Figure 8.  Deblurring results. (a)Original image, (b)blurred image generated by using k1 kernel located at right-bottom, (c)our proposed method, PSNR = 21.87dB, MSSIM = 0.94, (d)Goldstein and Fattal's method, PSNR = 21.50dB, MSSIM = 0.81, and (e)Krishnan et al.'s method, PSNR = 19.88dB, MSSIM = 0.70.
    Figure 9.  Deblurring results. (a)Original image, (b)blurred image generated by using k8 kernel located at right-bottom, (c)our proposed method, PSNR = 23.65dB, MSSIM = 0.95, (d)Goldstein and Fattal's method, PSNR = 23.39dB, MSSIM = 0.95, and (e)Krishnan et al.'s method, PSNR = 20.07dB, MSSIM = 0.89.
    Figure 10.  Deblurring results for real motion-blurred image (case 1). (a)Original image, (b)our proposed method, (c)Goldstein and Fattal's method, and (d)Krishnan et al.'s method.
    Figure 11.  Deblurring results for real motion-blurred image (case 2). (a)Original image, (b)our proposed method, (c)Goldstein and Fattal's method, and (d)Krishnan et al.'s method.

    The experimental results have been presented above. For computational time, our proposed method is superior to Goldstein and Fattal's method and Krishnan et al.'s method, as shown in Figure 7. For MSSIM and PSNR values, our proposed method is superior to Krishnan et al.'s method and is close to Goldstein and Fattal's method, illustrated in Table 1. For real blurry images, because many factors caused the motion-blurred images are unknown beforehand, the recovered image still existed the failure, as shown in Figure 11. In summary, the global performance for MSSIM and computational time is superior to that for the methods of Goldstein and Fattal and Krishnan et al.

    In this paper, we have proposed an image deblurring based on FPSF and clustering to recover the sharp image. Based on FPSF method applied the intelligence computing on the estimated image, the advantage of the computing strategy could reduce much more the computational complexity, at the same time, the optimal blur kernel could be estimated efficiently. As the experimental results, our proposed algorithm could efficiently reduce computational time. Also, for the blurry images, the proposed algorithm could restore the images with a good visual quality. It could enhance the image quality shown in many kinds of the view devices.

    There is no conflict of interest in this paper.



    [1] Dias C, Mendes L (2018) Protected Designation of Origin (PDO), Protected Geographical Indication (PGI) and Traditional Speciality Guaranteed (TSG): A bibliometric analysis. Food Res Int 103: 492–508. doi: 10.1016/j.foodres.2017.09.059
    [2] Litopoulou-Tzanetaki E, Tzanetakis N (2011) Microbiological characteristics of Greek traditional cheeses. Small Rum Res 101: 17–32. doi: 10.1016/j.smallrumres.2011.09.022
    [3] Montel MC, Buchin S, Mallet A, et al. (2014) Traditional cheeses: Rich and diverse microbiota with associated benefits. Int J Food Microbiol 177: 136–154. doi: 10.1016/j.ijfoodmicro.2014.02.019
    [4] Gatti M, Bottari B, Lazzi C, et al. (2014) Invited review: Microbial evolution in raw milk, long-ripened cheeses produced using undefined natural whey starters. J Dairy Sci 97: 573–591. doi: 10.3168/jds.2013-7187
    [5] Nacef M, Lelievre-Desmas M, Drider D, et al. (2019) Artisanal and industrial Maroilles cheeses: Are they different? Comparison using sensory, physico-chemical and microbiological approaches. Int Dairy J 89: 42–52.
    [6] Coelho MC, Silva CCG, Ribeiro SC, et al. (2014) Control of Listeria monocytogenes in fresh cheese using protective lactic acid bacteria. Int J Food Microbiol 191: 53–59. doi: 10.1016/j.ijfoodmicro.2014.08.029
    [7] Bautista-Gallego J, Alessandia V, Fontana M, et al. (2014) Diversity and functional characterization of Lactobacillus spp. isolated throughout the ripening of a hard cheese. Int J Food Microbiol 181: 60–66.
    [8] Ricciardi A, Guidone A, Ianniello RG, et al. (2015). A survey of non-starter lactic acid bacteria in traditional cheeses: Culture-dependent identification and survival to simulated gastrointestinal transit. Int Dairy J 43: 42–50. doi: 10.1016/j.idairyj.2014.11.006
    [9] Dolci P, Alessandria V, Rantsiou K, et al. (2008) Microbial dynamics of Castelmagno PDO, a traditional Italian cheese, with a focus on lactic acid bacteria ecology. Int J Food Microbiol 122: 302–311. doi: 10.1016/j.ijfoodmicro.2007.12.018
    [10] Domingos-Lopes MFP, Stanton C, Ross PR, et al. (2017) Genetic diversity, safety and technological characterization of lactic acid bacteria isolated from artisanal Pico cheese. Food Microbiol 63: 178–190. doi: 10.1016/j.fm.2016.11.014
    [11] Zoumpopoulou G, Tzouvanou A, Mavrogonatou E, et al. (2018) Probiotic features of lactic acid bacteria isolated from a diverse pool of traditional Greek dairy products regarding specific strain-host interactions. Prob Antimicrob Prot 10:313–322. doi: 10.1007/s12602-017-9311-9
    [12] Anonymous (2014) Cheeses of Protected Denomination of Origin. In: Hellenic Code of Food and Beverages, 3 Eds., 14–59.
    [13] Anonymous (1994) Recognition of a Protected Denomination of Origin (P.D.O.) for Galotyri cheese. In: Newspaper of the Government of the Republic of Greece, 51–62.
    [14] Rogga KJ, Samelis J, Kakouri A, et al. (2005) Survival of Listeria monocytogenes in Galotyri, a traditional Greek soft acid-curd cheese, stored aerobically at 4 and 12 ℃. Int Dairy J 15: 59–67. doi: 10.1016/j.idairyj.2004.05.002
    [15] Lekkas C, Kakouri A, Paleologos E, et al. (2006) Survival of Escherichia coli O157:H7 in Galotyri cheese stored at 4 and 12 ℃. Food Microbiol 23: 268–276. doi: 10.1016/j.fm.2005.03.008
    [16] Samelis J, Kakouri A (2007) Microbial and safety qualities of PDO Galotyri cheese manufactured at the industrial or artisan scale in Epirus, Greece. Ital J Food Sci 19: 91–99. doi: 10.1111/j.1365-2621.1984.tb00328.x
    [17] Pexara A, Solomakos N, Sergelidis D, et al. (2012). Fate of enterotoxigenic Staphylococcus aureus and staphylococcal enterotoxins in Feta and Galotyri cheeses. J Dairy Res 79: 405–413. doi: 10.1017/S0022029912000325
    [18] Katsiari MC, Kondyli E, Voutsinas LP (2009) The quality of Galotyri-type cheese made with different starter cultures. Food Control 20: 113–118. doi: 10.1016/j.foodcont.2008.02.011
    [19] Kondyli E, Katsiari MC, Voutsinas LP (2008) Chemical and sensory characteristics of Galotyri-type cheese made using different procedures. Food Control, 19: 301–307. doi: 10.1016/j.foodcont.2007.04.007
    [20] Kondyli E, Massouras T, Katsiari MC, et al. (2013) Lipolysis and volatile compounds of Galotyri-type cheese made using different procedures. Small Rum Res 113: 432–436. doi: 10.1016/j.smallrumres.2013.04.006
    [21] Kallinteri LD, Kostoula OK, Savvaidis IN (2013) Efficacy of nisin and/or natamycin to improve shelf-life of Galotyri cheese. Food Microbiol 36: 176–181. doi: 10.1016/j.fm.2013.05.006
    [22] AOAC (1995) AOAC official methods (16th. Edition) Methods 948.12, 935.43, 971.19.
    [23] Manolopoulou E, Sarantinopoulos P, Zoidou E, et al. (2003) Evolution of microbial populations during traditional Feta cheese manufacture and ripening. Int J Food Microbiol 82: 153–161. doi: 10.1016/S0168-1605(02)00258-1
    [24] Samelis J, Lianou A, Kakouri A, et al. (2009) Changes in the microbial composition of raw milk induced by thermization treatments applied prior to traditional Greek hard cheese processing. J Food Prot 72: 783–790. doi: 10.4315/0362-028X-72.4.783
    [25] Noutsopoulos D, Kakouri A, Kartezini E, et al. (2017) Growth, nisA gene expression and in situ nisin A activity of novel Lactococcus lactis subsp. cremoris costarter culture in commercial hard cheese production. J Food Prot 80: 2137–2146.
    [26] Salvetti E, Torriani S, Felis GE (2012) The genus Lactobacillus: A taxonomic update. Probiotics Antimicrob Prot 4: 217–226. doi: 10.1007/s12602-012-9117-8
    [27] Hammes WP, Hertel C (2009) Genus I Lactobacillus Beijernick 1901, 212 AL . In: Whitman WB, Bergey's Manual of Systematic Bacteriology, 2 Eds., New York: Springer, 465–511.
    [28] Holzapfel WH, Franz CMAP, Ludwig W, et al. (2009) Genus III Pediococcus Claussen 1903, 68AL. In: Whitman WB, Bergey's Manual of Systematic Bacteriology, The Firmicutes, 2 Eds., New York: Springer, 513–520.
    [29] Holzapfel WH, Bjorkroth JA, Dicks LMT (2009) Genus I Leuconostoc van Tieghem 1878, 198AL. In: Whitman WB, Bergey's Manual of Systematic Bacteriology, The Firmicutes, 2 Eds., New York: Springer, 624–635.
    [30] Manero A, Blanch AR (1999) Identification of Enterococcus spp. with a biochemical key. Appl Environ Microbiol 65: 4425–4430.
    [31] Vandera E, Kakouri A, Koukkou AI, et al. (2019) Major ecological shifts within the dominant nonstarter lactic acid bacteria in mature Greek Graviera cheese as affected by the starter culture type. Int J Food Microbiol 290: 15–26. doi: 10.1016/j.ijfoodmicro.2018.09.014
    [32] Harrigan WF, McCance ME (1976) Laboratory Methods in Food and Dairy Microbiology. Academic Press, New York.
    [33] Curk MC, Hubert JC, Bringel F (1996) Lactobacillus paraplantarum sp. nov., a new species related to Lactobacillus plantarum. Int J Syst Bacteriol 46: 595–598.
    [34] Alves PI, Martins MP, Semedo T, et al. (2004) Comparison of phenotypic and genotypic taxonomic methods for the identification of dairy enterococci. Ant Leeuwen 85: 237–252. doi: 10.1023/B:ANTO.0000020352.29496.4e
    [35] De Vos WM (1996) Metabolic engineering of sugar catabolism in lactic acid bacteria. Ant Leeuwen 70: 223–242. doi: 10.1007/BF00395934
    [36] Vaughan EE, van den Bogaard PTC, Catzeddu P, et al. (2001) Activation of silent gal genes in the lac-gal regulon of Streptococcus thermophilus. J Bacteriol 183: 1184–1194. doi: 10.1128/JB.183.4.1184-1194.2001
    [37] Salva S, Nunez M, Villena J, et al. (2011) Development of a fermented goat's milk containing Lactobacillus rhamnosus: In vivo study of health benefits. J Sci Food Agric 91: 2355–2362. doi: 10.1002/jsfa.4467
    [38] Olbrich dos Santos KM, Silva Vieira AD, Alonso Buriti FC, et al. (2015) Artisanal Coalho cheeses as a source of beneficial Lactobacillus plantarum and Lactobacillus rhamnosus strains. Dairy Sci Technol 95:209–230. doi: 10.1007/s13594-014-0201-6
    [39] Lo R, Ho VTT, Bansal N, et al. (2018) The genetic basis underlying variation in production of the flavour compound diacetyl by Lactobacillus rhamnosus strains in milk. Int J Food Microbiol 265: 30–39. doi: 10.1016/j.ijfoodmicro.2017.10.029
    [40] Innocente N, Biasutti M, Rita F, et al. (2016) Effect of indigenous Lactobacillus rhamnosus isolated from bovine milk on microbiological characteristics and aromatic profile of traditional yogurt. LWT Food Sci Technol 66: 158–164. doi: 10.1016/j.lwt.2015.10.031
    [41] Fernandez B, Vimont A, Desfosses-Foucault E, et al. (2017) Antifungal activity of lactic and propionic acid bacteria and their potential as protective culture in cottage cheese. Food Control 78: 350–356. doi: 10.1016/j.foodcont.2017.03.007
    [42] Gobbetti M, Morea M, Baruzzi F, et al. (2002) Microbiological, compositional, biochemical and textural characterisation of Caciocavallo Pugliese cheese during ripening. Int Dairy J 12: 511–523. doi: 10.1016/S0958-6946(02)00042-0
    [43] Matijasic BB, Rajsp MK, Perko B, et al. (2007) Inhibition of Clostridium tyrobutyricum in cheese by Lactobacillus gasseri. Int Dairy J 17: 157–166. doi: 10.1016/j.idairyj.2006.01.011
    [44] Arakawa K, Matsunaga K, Takihiro S, et al. (2015) Lactobacillus gasseri requires peptides, not proteins of free amino acids, for growth in milk. J Dairy Sci 98: 1593–1603. doi: 10.3168/jds.2014-8860
    [45] Kazou M, Alexandraki V, Blom J, et al. (2018) Comparative genomics of Lactobacillus acidipiscis ACA-DC 1533 isolated from traditional Greek Kopanisti cheese against species within the Lactobacillus salivarius clade. Front Microbiol 9: 1244. doi: 10.3389/fmicb.2018.01244
    [46] Litopoulou-Tzanetaki E, Tzanetakis N (2014) The microfloras of traditional Greek cheeses. Microbiol Spectrum 2: CM-0009-2012.
    [47] El-Baradei G, Delacroix-Buchet A, Ogier JC (2007) Biodiversity of bacterial ecosystems in traditional Egyptian Domiati cheese. Appl Environ Microbiol 73: 1248–1255. doi: 10.1128/AEM.01667-06
    [48] Barouei J, Karbassi A, Ghoddusi HB, et al. (2011) Impact of native Lactobacillus paracasei subsp. paracasei and Pediococcus spp. as adjunct cultures on sensory quality of Iranian white brined cheese. Int J Dairy Technol 64: 526–535.
    [49] Hatzikamari M, Litopoulou-Tzanetaki E, Tzanetakis N (1999) Microbiological characteristics of Anevato: a traditional Greek cheese. J Appl Microbiol 87: 595–601. doi: 10.1046/j.1365-2672.1999.00857.x
    [50] Xanthopoulos V, Polychroniadou A, Litopoulou-Tzanetaki E, et al. (2000) Characteristics of Anevato cheese made from raw or heat-treated goat milk inoculated with a lactic starter. Leben Wissen Technol 33: 483–488. doi: 10.1006/fstl.2000.0699
    [51] Samelis J, Delbes C, Kakouri A, et al. (2015) Differences in the biodiversity of lactic acid bacteria between an industrial-type and artisan-type Greek PDO Galotyri cheese. In: Book of Abstracts of the 7th IDF International Symposium on Sheep, Goat and Other Non-Cow Milk, 23–25 March 2015, Limassol, Cyprus, 26.
    [52] Parapouli M, Kakouri A, Koukkou AI, et al. (2015) Biochemical and molecular characterization of lactic acid bacteria isolated from two different types of Greek PDO Galotyri cheese. In: Book of Abstracts of the IAFP 2015 European Symposium on Food Safety, 20–22 April 2015, Cardiff, Wales, 80.
    [53] Beresford TP, Fitzsimons NA, Brennan NL, et al. (2001) Recent advances in cheese microbiology. Int Dairy J 11: 259–274. doi: 10.1016/S0958-6946(01)00056-5
    [54] Settanni L, Moschetti G (2010) Non-starter lactic acid bacteria used to improve cheese quality and provide health benefits. Food Microbiol 27: 691–697. doi: 10.1016/j.fm.2010.05.023
    [55] Gobbetti M, De Angelis M, Di Cagno, et al. (2015) Pros and cons for using non-starter lactic acid bacteria (NSLAB) as secondary/adjunct starters for cheese ripening. Trends Food Sci Technol 45: 167–178. doi: 10.1016/j.tifs.2015.07.016
  • Reader Comments
  • © 2019 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

Metrics

Article views(5256) PDF downloads(651) Cited by(12)

Figures and Tables

Figures(1)  /  Tables(9)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog