基于4-邻域的运动目标提取改进算法

Improved Algorithm of Extracting the Moving Object Based On the 4-Neighborhood

目标提取是图像模式识别的基本要素。传统的8-邻域目标提取算法可以提取目标,但是却不能区分不同的目标。在传统的8-邻域目标提取算法的基础上加以改进,采取基于4-邻域的目标提取算法,本算法通过4-邻域确定中心点像素的值,并同时能够区分目标,提取目标参数。通过实验显示,改进的算法能够提高目标识别的速度和精度,并能够改善系统的整体性能,在实际应用中取得较好的效果。

Target extraction is the basic elements of image pattern recognition,the current algorithm is based on 8-neighborhood sure pixels values,which the edge detection of target and obtain the objective basic outline.This kind of traditional methods can extract goal,but can not distinguish between different goals.Based on the traditional 8-neighborhood target extraction algorithm is improved,the basis of based on 4-neighborhood target extraction algorithm,this algorithm is determined by 4-neighborhood center,and also the value of pixels can discriminate between target,extracts parameters.Through the experiment shows that the improved algorithm can improve target recognition speed and precision,and can improve system performance and in the actual application of the good result are achieved.