基于热点区域的运动目标的检测

Detection of moving target based on hotspots

运动目标的检测是从图像序列中将运动目标所在的区域从背景当中分割出来。混合高斯模型是背景减除法中一种常用的背景更新算法,它能够完整地分割出运动目标所在的区域、适应规律性的噪音,但是对背景物体的突变无法及时处理,检测结果中会带有"伪前景"。对此,本文提出一种基于热点区域的运动目标的检测算法。实验结果表明该方法便捷快速,能够有效克服混合高斯模型的缺点。

Detecting moving targets from image sequences is to segment these targets from background. Gaussian mixture model in background subtraction method is a common background updating algorithm, it can completely segment the regions of moving targets and adapt to the regular noise. However, mutations in the background of the object cannot be processed in time, the results will be with a ＂pseudo prospects.＂ In this regard, this paper proposed a method based on hot spot regions to detect moving target. The experimental results show that this method is convenient and fast, and can effectively overcome the shortcomings of Gaussian mixture model.