论文部分内容阅读
本文提出的视频车速检测方法主要分为四个步骤:建立映射关系表、确定车尾、跟踪车尾、提取灰度直方图、计算车速。本文将提出一组利用视频图像计算出目标车辆任意位置瞬时车速的精确计算数学模型。根据目标检测的思想,帧差法能够较好的检测出运动目标位置,从每一个灰度直方图的最大值处按照该车速向前向后分别计算出各帧车尾在跟踪区域中所对应的位置,然后将8个灰度直方图中这些对应位置处的灰度值相加,作为本车速的扫描的最终结果。根据各个直方图的波形相似性和平移的车速相关性可知,扫描结果的最大值所对应的车速就是车辆的实际车速。
This paper presents a video speed detection method is mainly divided into four steps: the establishment of the mapping table, determine the rear, track the rear, extract the gray histogram, calculate the speed. In this paper, we propose a set of accurate mathematical models to calculate the instantaneous vehicle speed at any position of the target vehicle by using video images. According to the idea of target detection, the frame difference method can detect the moving target position well. From the maximum of each gray-scale histogram, the corresponding position of the rear-end of each frame in the tracking area is calculated according to the vehicle speed , And then add the gray values at the corresponding positions in the eight grayscale histograms as the final result of the scan of the own vehicle speed. According to the waveform similarity of each histogram and the translation speed correlation, the vehicle speed corresponding to the maximum value of the scanning result is the actual vehicle speed of the vehicle.