investigation of the noisy image edge detection based on the gwo algorithm

Edge detection as a pre-processing is the basis of image segmentation, feature extraction, and object recognition processes. so far, many edge detection algorithms have been introduced. however, even the best edge detectors lose their effectiveness in the presence of noise. therefore, the correct detection of edges in the noisy image is still one of the challenging issues in the image processing. various algorithms have been presented to solve this challenge, of which the meta-heuristic optimization algorithms are examples. in this research, a method is proposed for the edge detection of the noisy images based on the grey wolf optimization algorithm whose objective function is combining of the homogeneity factor, uniformity factor, and kirsch edge detector masks. the proposed method has been simulated on the bsds500 database including 500 images along with their ground truth images. in the simulation, two noises of the gaussian, and salt-and-pepper have been applied. the evaluation has been done according to the mean square error, peak signal-to-noise ratio, precision, f-score, and accuracy criteria. the simulation results show the mean accuracy of the proposed method on the bsds500 database images has achieved respectively 0.915, and 0.898 with the salt-and-pepper noise with a density of 0.01, and the gaussian noise with a zero mean, and a variance of 0.01. the average execution time of the proposed method with 80 runs for each image of the bsds500 database has also obtained at 50.01, and 50.02 seconds in the presence of the mentioned noises respectively.

investigation of the noisy image edge detection based on the gwo algorithm

edge detection as a pre-processing is the basis of image segmentation, feature extraction, and object recognition processes. so far, many edge detection algorithms have been introduced. however, even the best edge detectors lose their effectiveness in the presence of noise. therefore, the correct detection of edges in the noisy image is still one of the challenging issues in the image processing. various algorithms have been presented to solve this challenge, of which the meta-heuristic optimization algorithms are examples. in this research, a method is proposed for the edge detection of the noisy images based on the grey wolf optimization algorithm whose objective function is combining of the homogeneity factor, uniformity factor, and kirsch edge detector masks. the proposed method has been simulated on the bsds500 database including 500 images along with their ground truth images. in the simulation, two noises of the gaussian, and salt-and-pepper have been applied. the evaluation has been done according to the mean square error, peak signal-to-noise ratio, precision, f-score, and accuracy criteria. the simulation results show the mean accuracy of the proposed method on the bsds500 database images has achieved respectively 0.915, and 0.898 with the salt-and-pepper noise with a density of 0.01, and the gaussian noise with a zero mean, and a variance of 0.01. the average execution time of the proposed method with 80 runs for each image of the bsds500 database has also obtained at 50.01, and 50.02 seconds in the presence of the mentioned noises respectively.