论文部分内容阅读
针对目前遥感图像小目标检测算法遇到的复杂背景建模困难、先验信息严重匮乏等问题,考虑到昆虫视觉系统在图像处理方面的优势,提出基于蝇视觉系统大小场景(LF-SF)整合的信息处理模式解决遥感图像存在复杂背景的目标检测。蝇视觉的LF-SF整合机理无需考虑图像背景的复杂度以及目标先验信息,在抑制复杂背景纹理特征的同时对稀有目标特征进行增强,最后通过加权融合实现目标检测。实验结果表明,本文算法在算法设计、处理速度以及检测精度均优于现有算法。
Considering the advantages of insect visual system in image processing, aiming at the difficulty of complex background modeling and the shortage of a priori information in the small target detection algorithm for remote sensing images, the LF-SF integration based on flies vision system is proposed The information processing mode solves the target detection of remote sensing images with complex background. The LF-SF integration mechanism of fly vision does not need to consider the background complexity and target prior information, and enhances the rare target features while suppressing the complex background texture features. Finally, the target detection is achieved by weighted fusion. Experimental results show that the proposed algorithm is superior to the existing algorithms in algorithm design, processing speed and detection accuracy.