<p>Skin cancer, including melanoma, presents a serious health threat due to its potential to spread. Early detection remains essential for effective treatment. This paper introduces Velocity Hunger Games Search (VHGS), a multilevel thresholding method developed for segmenting skin cancer images. VHGS integrates a velocity-based optimization strategy with the Hunger Games Search algorithm to improve the exploration of complex solution spaces. The method is evaluated using two datasets: a benchmark set of 9 images and a dermoscopic dataset consisting of 2,500 skin cancer images. Its performance is assessed using Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Feature Similarity Indexing Method (FSIM). Two objective functions, Kapur’s entropy and Otsu’s between-class variance, guide the segmentation process. Experimental results demonstrate that VHGS consistently outperforms existing methods across various threshold levels. For example, in the second experiment, VHGS achieved the highest PSNR in 50% and 75% of the cases based on Kapur’s and Otsu’s methods, respectively. For FSIM, it ranked best in 50% and 87% of the evaluations using Kapur and Otsu. This study contributes to skin cancer diagnostics by presenting an effective multilevel thresholding algorithm capable of producing accurate and reliable segmentation for skin lesion image analysis.</p>