A Novel Radial Kernel Watershed Basis Segmentation Algorithm for Color Image Segmentation

Human vision cannot analyze the huge amount of information hidden in color images. Thus, amongst the research being developed currently, Color Image Segmentation (CIS) has become more vital. As each pixel in the image constitutes a different color complexion, which is tedious for estimating, CIS is a challenging process; also, it is a time-consuming process. Hence, this work proposes novel thresholding utilizing the Chaotic Logistic Fertile Field optimization Algorithm (CLFFA) approach and segmentation utilizing the Radial Kernel Watershed Basis Segmentation (RKWBS) technique for reducing time consumption. Utilizing the Contrast Limited Recursive Least Square Histogram Equalization (CLRLSHE) algorithm, the input image's contrast is enriched initially. Then, the Sobel edge detection detects the edges from the preprocessed image. Also, the CLFFA algorithm finds the optimal threshold value. After that, by thresholding the edge detected image grounded on RKWBS, the image is segmented quickly. Lastly, for removing the unwanted information in the segmented image, morphological operations are executed in it. Experiments were executed and analogized with the prevailing frameworks for proving the proposed mechanism's efficacy. The outcomes proved the efficacy regarding segmentation accuracy, computational time, et cetera.