强脉冲干扰下景物视频图像运动目标检测仿真  

Motion Target Detection Simulation of Scenery Video Image under Strong Pulse Interference

作　　者：胡明合[1] HU Ming-he(Computer engineering college,Shangqiu University ,Shangqiu Henan 476000,China)

机构地区：[1]商丘学院,计算机工程学院

出　　处：《计算机仿真》2019年第7期343-346,共4页Computer Simulation

摘　　要：图像处理技术作为当前应用范围较广的先进技术之一,对于社会各领域而言十分重要。针对当前视频图像运动目标检测存在的问题,提出基于运动信息的强脉冲干扰下景物视频图像运动目标检测方法。定义强脉冲干扰下景物视频图像集合、图像的结构元素和图像结构元素补集。根据设定参量分别对图像进行膨胀、腐蚀、开运算和闭运算等数学形态学操作,完成结构元素中突刺和强脉冲噪声的滤除,及目标图像中孔洞和间断的填补。根据视频图像像素点柯西分布概率的密度函数初步确定变化像素,分别定义低阈值检测出的所有变化区域和高阈值检测出的运动目标核心区域,得到候选阴影区域。利用判断规则确定图像阴影部分的像素点,并将其取出,获取运动目标,完成运动目标初步检测。给出图像局部统计中可变阈值,以此分割景物视频图像,控制杂散光导致的图像背景光强分布不均现象,同时通过灰度中心法对目标像素坐标点进行计算,实现视频图像运动目标精检测。实验结果表明,上述方法检测耗时短,检测性能强。In this paper, a method to detect the motion object of scene video image with strong impulse interference based on motion information was proposed. Based on strong pulse interference, we defined the set of scene video images, the structural elements of image and the complementary set of image structure elements. According to the setting parameters, we performed the mathematical morphology operations such as expansion, corrosion, opening and closing operations on the images, so as to complete the filtering of the thrust and strong impulse noise in the structural element and the filling of the holes and discontinuities in target image. According to the density function of Cauchy distribution probability of pixel in video image, we initially determined the variable pixels. Respectively, we defined all the changed regions detected by the low threshold and the core region of moving target detected by the high threshold, so as to obtain candidate shadow regions. Moreover, we used the judgment rule to determine the pixel point of shaded region in image and take out it, so as to obtain the moving target and complete the preliminary detection of moving target. In addition, we gave the variable threshold in the local statistics of image and thus to segment the scene video image. After that, we controlled the uneven distribution of background light intensity caused by the stray light. Meanwhile, we used the grayscale center method to calculate the target pixel coordinate point. Finally, we realized the fine detection for the moving target in video image. Simulation results show that the proposed method has short detection time and strong detection performance.

关 键 词：强脉冲 视频图像 目标 检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]