嵌入式运动目标提取实时图像处理系统

Design of a moving object real time image processing based on embedded system

针对大数据量、高速传输复杂的图像处理算法,以Matrox Oddysey图像处理卡为基础,设计了多目标实时图像处理系统。提出了一种结合帧间减法与背景自适应更新的算法,完成实时交会处理任务,满足靶场弹道试验任务的实时性要求。自适应背景更新是一种有效的用于视频序列图像运动目标检测的算法,该领域中很多方法被提出来,但是很少提及用于实时处理系统。我们着眼于算法的高速性并提出一种足够快的方法用于弹道测量系统。用高斯分布建立每个像素的灰度模型,这个分布用来区分前景和背景像素以便用来更新背景模型。

A Aiming at the needs of large data, high transport and complex arithmetic, a real time and multi-target imaging processing system is designed. A algorithm base on two consecutive frames and adaptive background updating is presented to do the real time intersection processing. It is satisfied real time demand of ballistic trajectory examination in shooting range. Adaptive background up- dating is one of the methods used to detect moving objects in video sequences. Many techniques have been presented in this field there are few of these methods is used in real-time application. We concentrated in the speed of algorithm and present a method that fast enough to be used in ballistic measuring system. We model the history of each pixel using Gaussian distribution. The distri- bution are used to classify the pixel to foreground and background pixel and thus to update the background model.