A Modified Snake Optimizer Algorithm with Otsu-based Method for Satellite Image Segmentation
Jiahao Fu, Rachsuda Setthawong · 2023
Image segmentation is an important step in image analysis that aims to segment regions of interest in an image by assigning a label to individual pixels sharing certain characteristics. Otsu-based method is a well-known thresholding technique that selects a threshold to segment regions by maximizing the variance between classes. Despite its advantages of considerable effectiveness and stability, its major drawback is high computational cost. This paper proposes a Modified Snake Optimizer algorithm (MSO), which can dynamically and efficiently tune Snake Optimizer (SO) parameters. To address the aforementioned drawback, MSO is applied with the Otsu threshold method (MSO-Otsu) in segmenting satellite images which helps analyze the snow-covered areas of mountain ranges in China. The experimental results show that the proposed MSO, in general, outperformed the traditional SO when applying to benchmark functions, and the proposed MSO-Otsu outperforms the traditional Otsu-based method in segmentation results and convergence time.