基于2维Tsallis熵的水下图像目标检测

Detection of Objects in Underwater Images Based on the Two-Dimensional Tsallis Entropy

针对传统图像检测方法在水下图像处理过程中存在目标区域定位不准确、目标细节丢失、目标形状变形的问题,文中利用Tsallis熵的非广延性,提出了一种基于边缘信息的2维直方图,并以最大2维Tsallis熵为准则,利用改进粒子群优化算法寻找最佳阈值.水下图像处理试验表明,该算法是一种有效的水下图像目标检测方法,与传统方法相比,具有更强的自适应性和鲁棒性.

For the problems in underwater image processing by traditional image detection methods,such as inaccurate location of objects regions,loss of object details and distortion of object shape,etc.,a new two-dimensional histogram based on edge information is proposed by utilizing the non-extensive property of Tsallis entropy.The improved particle swarm optimization（PSO） is used to search the best threshold value by maximizing the two-dimensional Tsallis entropy.The test results of some underwater images show that it is efficient to detect objects in underwater images.Comparing with traditional methods,the proposed approach shows better adaptability and robustness.