Multi-Threshold Image Segmentation for Biotechnological Applications Using an Improved Whale Optimization Algorithm

Abstract

Accurate and efficient image segmentation plays a critical role in various biotechnological applications, including medical imaging, cellular analysis, and bioinformatics, where high precision is essential for research and commercialization. Traditional multi-threshold image segmentation methods often face challenges such as high computational complexity and low segmentation efficiency. To address these issues, this study proposes a multi-threshold image segmentation algorithm based on an improved Whale Optimization Algorithm (WOA), tailored for biotechnological data analysis. The proposed approach enhances population diversity through Tent chaotic map optimization during the initialization phase. Additionally, it employs a nonlinear time-varying factor balance algorithm to improve global and local search capabilities. To further prevent the algorithm from falling into local optima, a sparrow watcher mechanism is integrated, facilitating population disturbance and enhancing inter-population communication. The algorithm's performance was quantitatively assessed using objective function values, peak signal-to-noise ratio (PSNR), standard deviation, and computational time in multi-threshold image segmentation tasks based on inter-class variance. Experimental results demonstrate significant improvements in convergence speed, segmentation accuracy, and algorithm stability compared to existing methods. The proposed algorithm offers a robust and efficient tool for processing complex biological images, providing a scalable solution that can support the commercialization of biotechnological innovations by improving diagnostic precision, accelerating research processes, and enhancing data-driven decision-making in the biotechnology sector.