This paper proposes a fast image-segmentation algorithm with a multilevel threshold based on the Tsallis relative entropy and wind-driven optimization algorithm. First, the principle of the Tsallis relative entropy is analyzed, and single threshold segmentation is extended to multilevel threshold segmentation. Then, a Gauss distribution is used to fit the image histogram information after segmentation, and the Tsallis relative entropy is used to determine the best segmentation threshold. To improve the speed of the threshold-segmentation algorithm, a wind-driven optimization algorithm is used to find the optimal solution of the Tsallis relative-entropy function. Finally, the proposed algorithm is compared with exhaustive and particle swarm optimization algorithms. The proposed algorithm is also compared with the Otsu algorithm and the multi threshold-segmentation method based on two-dimensional entropy. The experimental results show that the proposed algorithm can be used for multi-threshold segmentation of images with high speed and high accuracy.