摘要： 针对户外监视应用背景,提出了彩色图像差值模型、自适应阈值分割算法、图像形态学噪声滤除方法,实现了户外场景中运动目标的有效检测;对分割后的目标,提取不变矩特征,然后利用人工神经网络实现了运动目标的快速识别。实验结果验证了上述方法的有效性。

关键词: 运动目标检测, 视频监视系统, 图像分割, 人工神经网络, 图像形态学

Abstract: This paper first presents an efficient moving object detection method for outdoor surveillance environments. The presented method consists of a color image difference model, an adaptive thresholding method, and an image morphology filtering method. Moreover, by extracting shape invariant moment characteristics of object region, this paper also presents a BP neural network based object recognition method. Experimental results demonstrate that the proposed approach can detect and identify moving objects effectively for outdoor surveillance environments.

中图分类号: 

• TP391.41