视频桩考仪中的图像处理技术

设计了一种不需使用特定考车、可由计算机自动判定考试成绩的汽车驾驶员视频桩考仪。采用CCD摄像机和带有红外探测器的标志杆对考场进行实时监测。通过图像去噪、滤波的图像预处理、运动目标检测、目标跟踪算法、运动目标轨迹描述等图像处理技术,实现了由计算机实时对考车序列图像进行运动轨迹跟踪、桩考汽车的路线、方向、停车和中途停车等信息实时检测和考试人员的操作是否正确的自动判断,保证了考试的公平性和合理性。

Image Processing Techniques in Driver Bar-testing System

A driver bar-testing system of video for the drivers is presented which need not use special examination automobiles and examination results of drivers can be automatic confirmed. Using CCD and the sign pole with infrared detector, the system can achieve real time inspecting of examination field. By image processing techniques of image pre-processing of noise removing and filtering, moving target detecting, target tracking and trace tracking, it can fulfill real time moving trace tracking, automatic detection of trace, direction, stop and interval stop of the automobiles at image sequence and confirmation of legitimacy of drivers operation by computer. It can assure the equitableness and rationality.