摘要：

为提高公交系统智能化管理水平,提出一种公交客流自动统计方法。采用直方图统计与多帧平均的方法提取视频背景,使用背景边缘去除算法得到乘客目标边缘轮廓信息。在此基础上,根据乘客头部轮廓的类圆特性, 利用基于梯度信息的Hough 变换圆检测算法完成乘客头部轮廓的识别,通过基于Kalman 滤波预测的CamShift 目标跟踪算法实现乘客的检测与计数。实验结果表明,该方法能有效消除背景图像中的噪声以及背景边缘,准确识别乘客目标并对其跟踪计数,从而提高城市公共交通运输效率。

关键词: 客流统计, Canny 边缘检测, Hough 变换, CamShift 目标跟踪, Kalman 滤波

Abstract:

In order to improve the intelligent management level of public transport system,this paper proposes an automatic passenger flow statistical method. It uses a background extraction algorithm based on statistics of histogram combining with multi-frame average to extract video background, and uses background edge removal algorithm to eliminate most of the background information. According to the circular-like feature of passenger head contour,it employs Hough transform based on gradient information detection and Camshift target tracking based on Kalman filtering prediction to complete passenger tracking and counting. Experimental results show that this method can effectively eliminate the background noise and background edges in the picture,identify the passengers target to track them and count accurately,and improve the efficiency of urban public transportation.

Key words: passenger flow statistics, Canny edge detection, Hough transform, Camshift target tracking, Kalman filtering

中图分类号: 

• TP391