Apparently, the aim of this paper is to segment the defective part from the defective image in the leather industry. The leather industry keeps on being the second biggest work producing division in India. In this proposed work, the input defective image from the leather industry is utilized then the preprocessing technique is employed for removing the noise in the image. For removing, the noise RGB images is takes place and convert to grey image after that the images are connected to segmentation part. In this segmentation process the Region Growing (RG) technique is carried out then for improving the image Modified Region Growing (MRG) segmentation is executed with seed point determination and thresholding. For optimizing, the threshold value three different optimization algorithms utilized such as Grey Wolf Optimization(GWO), Particle Swarm Optimization(PSO) Algorithm and Cuckoo Search (CS). The obtained optimal threshold is considered to get the defective part segmented. In the result, convergence graph for three algorithms is utilized and the performance matrices in terms of Accuracy, Sensitivity, Specificity, PPV, NPV, FPV and FDR are examined. The sensitivity value for GWO is 92%, CS is 85% and PSO is 80% based on these GWO is employed better compared with other two techniques.
Volume 11 | Issue 1
Pages: 274-284