基于视觉注意的光学遥感图像舰船目标检测方法

Ship detection algorithm in optical remote-sensing images based on visual attention

根据大视场光学遥感图像中舰船目标检测的实际需求,借鉴了人眼视觉注意感知的特性,引入一种基于自适应形态学的显著图快速生成方法。通过加入杂波率系数来自适应地控制Top-Hat算子中的结构元素尺寸,结合一组模拟人眼视觉感知特性的Gabor滤波器组,在有效抑制复杂背景的同时,实现了舰船目标的提取与检测。实验表明,该算法在各种海况下有较高的检测率与较低的虚警率,比传统算法的检测效果有较大改善。

According to the actual needs of the large field of optical remote sensing images of ship targets detection, referenced by the perceived characteristics of human visual attention, an adaptive morphological strategy was em- ployed to generate saliency map. A clutter rate coefficient is used to control the structural elements of the morpho- logical Top-Hat operator. Then, a set of Gabor filters are added to simulate human vision perception. Under reduc- ing the negative influence of complex background, they can extract and detect the ship target. The experiments show that the algorithm has a higher detection rate and lower false alarm rate in a variety of sea conditions, and it improves the detection results than the traditional algorithm.