With the rapid development of space technology, space background of the division of non-cooperative target has become the new focus of concern. SUSAN algorithm is a new class of parallel boundary segmentation algorithm using the USAN principle, covered by the template target pixel to extract the statistical features. Characteristics of the target image for the space, this paper presents a space-based object segmentation algorithm SUSAN algorithm. Target by the edges in the image information extraction, to achieve the artificial separation of target and background. The algorithm has good noise immunity, feature location accuracy, computing speed, better able to maintain the structural information of the characteristics of the image features, ideal for real-time image segmentation space.%随着航天科技的迅猛发展,空间背景下非合作目标的分割问题已经成为人们关注的新焦点.SUSAN算法是一种新兴的并行边界类分割算法,采用USAN原理,通过对模板覆盖像素的统计来提取目标的特征.针对空间目标图像的特点,提出了一种基于SUSAN算法的空间目标分割算法.利用图像中目标的边缘轮廓信息进行特征提取,实现了人造目标与背景的分离.该算法具有抗噪声能力好、特征定位准确、计算速度快、能够较好的保持图像的特征结构信息等特点,非常适用于航天图像的实时分割处理.
展开▼
