In the present study, we aim to propose an effective and robust ensemble-learning approach with s... more In the present study, we aim to propose an effective and robust ensemble-learning approach with stacked generalization for image segmentation. Initially, the input images are processed for feature extraction and edge detection using the Gabor filter and the Canny algorithms, respectively; our main goal is to determine the most feature descriptions. Subsequently, we applied the stacking generalization technique, which is generally built with two main learning levels. The first level is composed of two algorithms that give good results in the literature, namely: LightGBM (Light Gradient Boosting Machine) and SVM (support vector machine). The second level is the meta-model in which we use a predictor model that takes the base-level predictions to improve the accuracy of the final prediction. In the stacked generalization process, we use the Extreme Gradient Boosting (XGBoost); it takes as input the sub-models’ outputs to better classify each pixel of the image to give the final predict...
2020 Fourth International Conference On Intelligent Computing in Data Sciences (ICDS)
Image segmentation is the most important operation in the image processing system because it is l... more Image segmentation is the most important operation in the image processing system because it is located at the articulation between image processing and analysis. The advantage of segmentation is to partition an image into several homogeneous regions, within the meaning of a criterion fixed a priori. A multitude of segmentation methods are proposed in the literature, but there is no universal image segmentation technique to apply to all different types of images and in any given computer context. Because of these constraints, in this paper, we will propose a new image segmentation approach which is based on the hybridization of an unsupervised classification method which is fuzzy C-means (FCM) and a metaheuristic which is called Cuckoo Search Algorithm (CSA). In our proposed approach, the cuckoo search is used to find the optimal partitioning according to an objective function which is based on the indices of validity of the clusters. First, CSA is initialized with random cluster centers. The cluster centers are then updated using the CSA principles aimed at minimizing the objective function proposed. The performance of the proposed approach was measured on several images and compared to other existing FCM techniques such as standard FCM and FCM based on genetic algorithms (FCM-GA). The experimental results show that the proposed approach yields satisfactory results in terms of precision, simplicity and efficiency.
Image segmentation is a fundamental and important step in many computer vision applications. One ... more Image segmentation is a fundamental and important step in many computer vision applications. One of the most widely used image segmentation techniques is clustering. It is a process of segmenting the intensities of a non-homogeneous image into homogeneous regions based on their similarity property. However, clustering methods require a prior initialization of random clustering centers and often converge to the local optimum, thanks to the choices of the initial centers, which is a major drawback. Therefore, to overcome this problem, we used the improved version of the sine-cosine algorithm to optimize the traditional clustering techniques to improve the image segmentation results. The proposed method provides better exploration of the search space compared to the original SCA algorithm which only focuses on the best solution to generate a new solution. The proposed ISCA algorithm is able to speed up the convergence and avoid falling into local optima by introducing two mechanisms th...
Image segmentation is a fundamental and important step in many computer vision applications. One ... more Image segmentation is a fundamental and important step in many computer vision applications. One of the most widely used image segmentation techniques is clustering. It is a process of segmenting the intensities of a nonhomogeneous image into homogeneous regions based on their similarity property. However, clustering methods require a prior initialization of random clustering centers and often converge to the local optimum, thanks to the choices of the initial centers, which is a major drawback. Therefore, to overcome this problem, we used the improved version of the sine-cosine algorithm to optimize the traditional clustering techniques to improve the image segmentation results. The proposed method provides better exploration of the search space compared to the original SCA algorithm which only focuses on the best solution to generate a new solution. The proposed ISCA algorithm is able to speed up the convergence and avoid falling into local optima by introducing two mechanisms that take into account the first is the given random position of the search space and the second is the position of the best solution found so far to balance the exploration and exploitation. The performance of the proposed approach was evaluated by comparing several clustering algorithms based on metaheuristics such as the original SCA, genetic algorithms (GA) and particle swarm optimization (PSO). Our evaluation results were analyzed based on the best fitness values of several metrics used in this paper, which demonstrates the high performance of the proposed approach that gives satisfactory results compared to other comparison methods.
2019 7th Mediterranean Congress of Telecommunications (CMT)
Image segmentation is an important step in any image analysis process. In the literature, there a... more Image segmentation is an important step in any image analysis process. In the literature, there are two dual approaches. The contour segmentation approach consists in locating the object boundaries and the segmentation approach by region consists in partitioning the image into a set of regions. The best results of segmentation are obtained by cooperating these two approaches. They are more effective because the disadvantages of one method can be overcome by the advantages of another method. In an image processing system, the most important operation is image segmentation. To date, there is no universal method of image segmentation. Any technique is only effective for a given type of image, for a given type of application, and in a given computer context. Because of these constraints, the various image processing strategies that have been proposed have asserted their inadequacies and limitations. It is therefore perfectly normal to explore new horizons and find new methods that are more flexible and more effective. In this paper, we will propose and discuss a cooperative approach between a k-means unsupervised classification method and Canny's contour detection method to improve color image segmentation results. First, we will apply the contour segmentation process by Canny's algorithm, and then we integrate the results obtained by another unsupervised classification segmentation process, which is k-means. Experimental results show the strengths of our approach in terms of accuracy, simplicity and efficiency.
2020 Fourth International Conference On Intelligent Computing in Data Sciences (ICDS), 2020
Image segmentation is the most important operation in the image processing system because it is l... more Image segmentation is the most important operation in the image processing system because it is located at the articulation between image processing and analysis. The advantage of segmentation is to partition an image into several homogeneous regions, within the meaning of a criterion fixed a priori. A multitude of segmentation methods are proposed in the literature, but there is no universal image segmentation technique to apply to all different types of images and in any given computer context. Because of these constraints, in this paper, we will propose a new image segmentation approach which is based on the hybridization of an unsupervised classification method which is fuzzy C-means (FCM) and a metaheuristic which is called Cuckoo Search Algorithm (CSA). In our proposed approach, the cuckoo search is used to find the optimal partitioning according to an objective function which is based on the indices of validity of the clusters. First, CSA is initialized with random cluster ce...
Segmentation is the heart of an automatic image analysis system. There are several segmentation t... more Segmentation is the heart of an automatic image analysis system. There are several segmentation techniques. The contour segmentation approach aims to separate regions of different gray levels and relatively homogeneous. The region segmentation approach consists in grouping the adjacent pixels of the image into distinct regions. The cooperative approach is used in order to improve the result of segmentation and the segmentation based on classification methods; in this case, classes are defined by the maximum sets of related pixels belonging to the same class. In this work by studying the dual problem, we develop a simple but efficient cooperative approach between a Random forest classification method and a set of contour detection methods as Canny, Prewitt and Sobel. Firstly, original image is initially segmented by hybridization of Canny, Prewitt, and Sobel’s algorithms for edge detection. Then, we will use the output image obtained by another supervised classification segmentation ...
This article presents a new image segmentation approach based on the principle of clustering opti... more This article presents a new image segmentation approach based on the principle of clustering optimized by the metaheuristic algorithm namely: SCA (Algorithm Sinus Cosine). This algorithm uses a mathematical model based on trigonometric functions to solve optimization problems. Such an approach was developed to solve the drawbacks existing in classic clustering techniques such as the initialization of cluster centers and convergence towards the local optimum. In fact, to obtain an "optimal" cluster center and to improve the image segmentation quality, we propose this technique which begins with the generation of a random population. Then, we determine the number of clusters to exploit. Later, we formulate an objective function to maximize the interclass distance and minimize the intra-class distance. The resolution of this function gives the best overall solution used to update the rest of the population. The performances of the proposed approach are evaluated using a set of reference images and compared to several classic clustering methods, like k-means or fuzzy c-means and other meta-heuristic approaches, such as genetic algorithms and particle swarm optimization. The results obtained from the different methods are analyzed based on the best fitness values, PSNR, RMSE, SC, XB, PC, S, SC, CE and the computation time. The experimental results show that the proposed approach gives satisfactory results compared to the other methods.
In the present study, we aim to propose an effective and robust ensemble-learning approach with s... more In the present study, we aim to propose an effective and robust ensemble-learning approach with stacked generalization for image segmentation. Initially, the input images are processed for feature extraction and edge detection using the Gabor filter and the Canny algorithms, respectively; our main goal is to determine the most feature descriptions. Subsequently, we applied the stacking generalization technique, which is generally built with two main learning levels. The first level is composed of two algorithms that give good results in the literature, namely: LightGBM (Light Gradient Boosting Machine) and SVM (support vector machine). The second level is the meta-model in which we use a predictor model that takes the base-level predictions to improve the accuracy of the final prediction. In the stacked generalization process, we use the Extreme Gradient Boosting (XGBoost); it takes as input the sub-models’ outputs to better classify each pixel of the image to give the final predict...
2020 Fourth International Conference On Intelligent Computing in Data Sciences (ICDS)
Image segmentation is the most important operation in the image processing system because it is l... more Image segmentation is the most important operation in the image processing system because it is located at the articulation between image processing and analysis. The advantage of segmentation is to partition an image into several homogeneous regions, within the meaning of a criterion fixed a priori. A multitude of segmentation methods are proposed in the literature, but there is no universal image segmentation technique to apply to all different types of images and in any given computer context. Because of these constraints, in this paper, we will propose a new image segmentation approach which is based on the hybridization of an unsupervised classification method which is fuzzy C-means (FCM) and a metaheuristic which is called Cuckoo Search Algorithm (CSA). In our proposed approach, the cuckoo search is used to find the optimal partitioning according to an objective function which is based on the indices of validity of the clusters. First, CSA is initialized with random cluster centers. The cluster centers are then updated using the CSA principles aimed at minimizing the objective function proposed. The performance of the proposed approach was measured on several images and compared to other existing FCM techniques such as standard FCM and FCM based on genetic algorithms (FCM-GA). The experimental results show that the proposed approach yields satisfactory results in terms of precision, simplicity and efficiency.
Image segmentation is a fundamental and important step in many computer vision applications. One ... more Image segmentation is a fundamental and important step in many computer vision applications. One of the most widely used image segmentation techniques is clustering. It is a process of segmenting the intensities of a non-homogeneous image into homogeneous regions based on their similarity property. However, clustering methods require a prior initialization of random clustering centers and often converge to the local optimum, thanks to the choices of the initial centers, which is a major drawback. Therefore, to overcome this problem, we used the improved version of the sine-cosine algorithm to optimize the traditional clustering techniques to improve the image segmentation results. The proposed method provides better exploration of the search space compared to the original SCA algorithm which only focuses on the best solution to generate a new solution. The proposed ISCA algorithm is able to speed up the convergence and avoid falling into local optima by introducing two mechanisms th...
Image segmentation is a fundamental and important step in many computer vision applications. One ... more Image segmentation is a fundamental and important step in many computer vision applications. One of the most widely used image segmentation techniques is clustering. It is a process of segmenting the intensities of a nonhomogeneous image into homogeneous regions based on their similarity property. However, clustering methods require a prior initialization of random clustering centers and often converge to the local optimum, thanks to the choices of the initial centers, which is a major drawback. Therefore, to overcome this problem, we used the improved version of the sine-cosine algorithm to optimize the traditional clustering techniques to improve the image segmentation results. The proposed method provides better exploration of the search space compared to the original SCA algorithm which only focuses on the best solution to generate a new solution. The proposed ISCA algorithm is able to speed up the convergence and avoid falling into local optima by introducing two mechanisms that take into account the first is the given random position of the search space and the second is the position of the best solution found so far to balance the exploration and exploitation. The performance of the proposed approach was evaluated by comparing several clustering algorithms based on metaheuristics such as the original SCA, genetic algorithms (GA) and particle swarm optimization (PSO). Our evaluation results were analyzed based on the best fitness values of several metrics used in this paper, which demonstrates the high performance of the proposed approach that gives satisfactory results compared to other comparison methods.
2019 7th Mediterranean Congress of Telecommunications (CMT)
Image segmentation is an important step in any image analysis process. In the literature, there a... more Image segmentation is an important step in any image analysis process. In the literature, there are two dual approaches. The contour segmentation approach consists in locating the object boundaries and the segmentation approach by region consists in partitioning the image into a set of regions. The best results of segmentation are obtained by cooperating these two approaches. They are more effective because the disadvantages of one method can be overcome by the advantages of another method. In an image processing system, the most important operation is image segmentation. To date, there is no universal method of image segmentation. Any technique is only effective for a given type of image, for a given type of application, and in a given computer context. Because of these constraints, the various image processing strategies that have been proposed have asserted their inadequacies and limitations. It is therefore perfectly normal to explore new horizons and find new methods that are more flexible and more effective. In this paper, we will propose and discuss a cooperative approach between a k-means unsupervised classification method and Canny's contour detection method to improve color image segmentation results. First, we will apply the contour segmentation process by Canny's algorithm, and then we integrate the results obtained by another unsupervised classification segmentation process, which is k-means. Experimental results show the strengths of our approach in terms of accuracy, simplicity and efficiency.
2020 Fourth International Conference On Intelligent Computing in Data Sciences (ICDS), 2020
Image segmentation is the most important operation in the image processing system because it is l... more Image segmentation is the most important operation in the image processing system because it is located at the articulation between image processing and analysis. The advantage of segmentation is to partition an image into several homogeneous regions, within the meaning of a criterion fixed a priori. A multitude of segmentation methods are proposed in the literature, but there is no universal image segmentation technique to apply to all different types of images and in any given computer context. Because of these constraints, in this paper, we will propose a new image segmentation approach which is based on the hybridization of an unsupervised classification method which is fuzzy C-means (FCM) and a metaheuristic which is called Cuckoo Search Algorithm (CSA). In our proposed approach, the cuckoo search is used to find the optimal partitioning according to an objective function which is based on the indices of validity of the clusters. First, CSA is initialized with random cluster ce...
Segmentation is the heart of an automatic image analysis system. There are several segmentation t... more Segmentation is the heart of an automatic image analysis system. There are several segmentation techniques. The contour segmentation approach aims to separate regions of different gray levels and relatively homogeneous. The region segmentation approach consists in grouping the adjacent pixels of the image into distinct regions. The cooperative approach is used in order to improve the result of segmentation and the segmentation based on classification methods; in this case, classes are defined by the maximum sets of related pixels belonging to the same class. In this work by studying the dual problem, we develop a simple but efficient cooperative approach between a Random forest classification method and a set of contour detection methods as Canny, Prewitt and Sobel. Firstly, original image is initially segmented by hybridization of Canny, Prewitt, and Sobel’s algorithms for edge detection. Then, we will use the output image obtained by another supervised classification segmentation ...
This article presents a new image segmentation approach based on the principle of clustering opti... more This article presents a new image segmentation approach based on the principle of clustering optimized by the metaheuristic algorithm namely: SCA (Algorithm Sinus Cosine). This algorithm uses a mathematical model based on trigonometric functions to solve optimization problems. Such an approach was developed to solve the drawbacks existing in classic clustering techniques such as the initialization of cluster centers and convergence towards the local optimum. In fact, to obtain an "optimal" cluster center and to improve the image segmentation quality, we propose this technique which begins with the generation of a random population. Then, we determine the number of clusters to exploit. Later, we formulate an objective function to maximize the interclass distance and minimize the intra-class distance. The resolution of this function gives the best overall solution used to update the rest of the population. The performances of the proposed approach are evaluated using a set of reference images and compared to several classic clustering methods, like k-means or fuzzy c-means and other meta-heuristic approaches, such as genetic algorithms and particle swarm optimization. The results obtained from the different methods are analyzed based on the best fitness values, PSNR, RMSE, SC, XB, PC, S, SC, CE and the computation time. The experimental results show that the proposed approach gives satisfactory results compared to the other methods.
Uploads
Papers by Lahbib Khrissi