基于矩阵像素阈值判定的运动目标检测

Moving object detection based on pixel matrix

针对运动目标检测所存在的不能完整分割运动对象、对光照和外部条件造成的场景变化比较敏感、容易出现许多伪运动目标点等情况,以背景差分算法为基础,提出了一种基于矩阵像素阈值判定的运动目标检测算法.该算法通过对图像进行预处理,增强图像中目标和背景图像的对比度,并去除采集和传输图像时产生的噪声,再将每个像素点的矩阵进行差分运算,以确定像素灰度变化的情况,利用灰度变化阈值确定最终的运动目标.与现有的背景差分算法检测到的运动目标相比,该算法的准确度更高.

It＇s often hard to cut up moving objects completely in the motion detection,which is prone to emerge many false moving points and sensitive to the scene changes caused by light and external conditions.In view of this situation,a motion detection algorithm which based on background subtraction algorithm and used matrix pixel threshold determination was proposed.This algorithm firstly preprocesses the images to enhance the contrast of the object and background images,and remove the noise generated by the collection and transmission of images,then conduct subtraction operation with the matrixes of each pixel line by line to determine the condition of pixels,thus using the gray changing threshold to determine the final moving objects.Compared with the moving objects detected by the existing background subtraction algorithms,this algorithm shows higher accuracy.