基于人类视觉机制的红外目标检测方法

摘要

针对红外小目标图像信噪比低,背景复杂等特点,提出一种基于人类视觉系统对比机制的红外小目标检测方法.该方法充分模拟人眼视觉系统对比机制,通过图像局部视觉对比度显著图及自适应阈值选取,锁定感兴趣目标区域,实现对红外图像中弱小目标的精确检测.与Top-hat和逐个像素对比度计算的检测方法进行了对比,实验结果表明所提出的方法能够在低信噪比条件下更有效地检测红外弱小目标.%For signal to noise ratio(SNR)of infrared small target image is low,and the background is complex and so on,an infrared small target detection method based on the contrast mechanism of human visual system is proposed.The method fully simulated the contrast mechanism of human visual system.The target region is locked by image local visual contrast saliency map and adaptive threshold selection to realize the accurate detection of small target in infrared image sequences.The method is compared with Top-hat and pixel by pixel contrast calculation method.The experimental results show that this method can detect the infrared dim target more effectively in low SNR condition.