Image Segmentation Techniques: Partitioning Images into Characteristic Regions and Target Extraction

Image segmentation refers to the technical process of dividing images into regions with distinct characteristics and extracting regions of interest, serving as a critical transition from image processing to image analysis. In wheat image processing, segmentation algorithms are employed to extract valuable image information, enabling the separation of adhered wheat kernels and providing foundational data for quality analysis through image processing techniques. This study categorizes segmentation algorithms and examines their fundamental principles, specifically investigating wheat image segmentation methods under varying lighting conditions, backgrounds, and connectivity scenarios. Key focus areas include wavelet transform-based watershed algorithms and threshold segmentation techniques, implemented and evaluated using MATLAB for adhesive wheat kernel separation.