基于HSV颜色空间的运动目标检测方法

摘要

针对在目标检测过程中光照变化对检测结果的影响，提出一种基于HSV颜色空间的运动目标检测方法。首先采用帧间差分分离出运动目标区域，将该区域转换为HSV颜色空间，用形态学对区域进行处理，消除噪声影响。然后利用各分量的相互独立性及H分量对光照不敏感的特性，应用自适应阈值分割方法，对运动目标实现准确分割。实验表明，该方法能够实现对运动目标的准确检测，消除帧间差分产生的图像空洞，对光照变化剧烈的目标也能实现准确分割，具有较强的鲁棒性。%　　Aimed at the influence of illumination changes on the test results in the process of target detection,a moving tar⁃get detection method based on HSV color space is presented. The inter⁃frame difference method is adopted to find out moving tar⁃get area,to convert the area to HSV color space,and to deal with it by morphology to eliminate the noise impact. By means of the independence of each component and the characteristics of H component insensitivity to light,the adaptive threshold spli⁃tting methods is adopted to achieve accurate segmentation of the moving target. The experiments show that the method can achieve accurate detection of moving targets,and eliminate the image empty generated by inter⁃frame difference,achieve accu⁃rate segmentation of target with the light variations,and has strong robustness.