混合背景下像素均值运动目标特征提取算法

Feature Extraction Algorithm for Moving Target with Pixel Mean in Hybrid Differential Background

针对传统方法在混合差分背景下对运动目标检测准确性不好的问题,为了提高运动目标检测识别能力,提出一种基于混合背景差法和像素均值技术的运动目标提取算法。构建运动目标的三维成像模型,对运动目标图像采用混合差分背景分割方法进行图像的模板匹配和自适应分割处理,结合几何边缘重构方法进行运动目标的像素特征提取,对运动图像使用混合差分背景分割方法进行图像的模板匹配和自适应分割处理;采用角点检测方法进行运动目标图像纹理渲染,对运动目标图像的像素信息采用帧分解和像素均值技术进行运动目标的特征提取;结合二值化处理技术进行运动目标的边缘轮廓检测和特征搜索,实现混合差分背景下的像素均值运动目标特征提取。仿真结果表明,采用该算法进行运动目标特征提取的图像处理能力较好,输出图像质量较高,特征提取的准确性较好,提高了运动目标的特征提取和检测能力。

In order to solve the problem of poor accuracy of traditional methods for moving target detection in hybrid differential background,and improve the ability of moving target detection,a moving target extraction algorithm based on mixed background difference method and pixel mean technique is proposed.The 3D imaging model of moving object is constructed,and the image template matching and adaptive segmentation are processed by using mixed differential background segmentation method,and the pixel feature extraction of moving object is carried out by combining geometric edge reconstruction method.The mixed differential background segmentation method is used for image template matching and adaptive segmentation,and corner detection method is used for texture rendering of moving object image.The pixel information of moving object image is extracted by frame decomposition and pixel mean technology,and the edge contour detection and feature search of moving object are carried out by combining binary processing technology.The feature extraction of moving target with pixel mean value in mixed differential background is realized.The simulation shows that the algorithm has better image processing ability,higher output image quality and better accuracy of feature extraction,and improves the ability of feature extraction and detection of moving targets.