论文部分内容阅读
在智能交通系统中,动态图像识别技术是系统应用的基础核心技术之一。以应用于交通监控、智能驾驶系统等场景的HSV空间动态车辆识别为基础,研究并论证提出了新的检测识别方法,实现对运动车辆的检测识别、目标追踪、驾驶辅助等功能。研究问题的难点是,如何从复杂的背景中分割运动物体,是检测方法能否有效的至关重要的一步,在研究了目前存在的各种方法之后,提出了一种新的基于阴影检测的HSV空间自适应背景模型的车辆追踪检测算法,算法基于HSV空间图像处理,采用最大类间方差法获取相邻帧二值化阈值,利用纹理信息进一步确定动态图像以及确认图像范围。通过截取由监控系统获取的视频信息,并对其进行图像处理检测车辆移动轨迹。从监控视频信息中获取两帧不同时刻的图像信息,在HSV空间进行相邻帧检测。由于阈值的选择将直接影响判断精度,本研究将固定阈值法进行了改进,该阈值是通过统计模型对整幅图像上灰度值进行计算,并通过最大类间方差法确定阈值。最后经过实际视频图像验证,仿真试验流程清晰,试验结果达到预期设想。
In the intelligent transportation system, the dynamic image recognition technology is one of the basic core technologies of system application. Based on HSV spatial dynamic vehicle identification applied to traffic monitoring and intelligent driving system, a new detection and recognition method is proposed and demonstrated to realize the functions of detection and recognition of moving vehicles, target tracking and driving assistance. The difficulty of the research problem is how to segment the moving objects from the complex background is a crucial step for the detection method to be effective. After studying various existing methods, a new shadow detection based HSV spatial adaptive background model. The algorithm is based on HSV spatial image processing, adopts maximum inter-class variance method to obtain the threshold value of adjacent frames, uses the texture information to further determine the dynamic image and confirm the image range. By intercepting the video information obtained by the monitoring system, and image processing to detect the vehicle trajectory. From the monitoring video information, two frames of image information are acquired at different times, and adjacent frames are detected in the HSV space. Since the selection of the threshold will directly affect the accuracy of the judgment, this study improves the fixed threshold method, which uses the statistical model to calculate the gray value of the entire image and determines the threshold by the maximum inter-class variance. Finally, after the actual video image verification, simulation test flow is clear, the test results reach the expected scenario.