MORPHOLOGICAL GRADIENT-BASED WATERSHED ALGORITHM FOR COLOR IMAGE SEGMENTATION

Abstract

Image segmentation and its performance evaluation are very difficult but important problems in computer vision. A major challenge The conventional watershed algorithm suffers from over-segmentation and is affected by light reflections in an image. We propose an enhanced watershed color image segmentation algorithm. It is founded on a morphological gradient. This method acquires the component gradient of a color image in a new color space that is unaffected by reflected light. The gradient image is reconstructed through the processes of opening and closing. Consequently, the ultimate gradient image is acquired. The maximum inter-class variance algorithm is employed to automatically determine the threshold for the final gradient image. The original gradient image is forcibly aligned with the acquired binary labeled image, and the adjusted gradient image is segmented using the watershed method. Experimental findings indicate that the proposed method can achieve an accurate and continuous target contour. It will attain the minimum requisite number of segmentation regions in accordance with human vision. In comparison to analogous algorithms, this method can mitigate the extraneous regions produced by reflected light. It will effectively preserve the object's edge information. It will enhance the robustness and applicability. The experimental results indicate that the proposed algorithm demonstrates a significant enhancement in operational efficiency, surpassing the region-growing method and the automatic threshold method by 10%. The proposed algorithm exhibits an accuracy and recall rate exceeding 0.98. The experimental comparison clearly demonstrates the advantages of the proposed algorithm in object segmentation.