基于HALCON的交通路口车流检测方法研究

 

 

(西安公路研究院,陕西 西安710065

  要:本文针对智能交通中的车流检测问题,利用Halcon计算机视觉算法库设计了交通路口车流视频图像处理检测算法,并对基于混合高斯背景模型的运动目标检测方法进行了应用研究与对算法基本原理进行了分析,选取Halcon中的关键函数并在HDevelop开发环境中实现了检测算法。在验证阶段,对连续500帧图像进行了处理分析,试验结果说明算法能够较为有效地提取路面运动车辆目标,平均每帧处理速度为5.8ms,能够满足实时性要求。

关键词:智能交通;车流检测;机器视觉;Halcon

中图分类号:U495        文献标识码:J

 

Research on the Method of Traffic Flow Detection at Traffic Junction Based on Halcon

JIAO Ming

(Xi’an Highway ResearchInstitute, Xi’an 710065,China)

Abstract: In this paper, a video image processing detection algorithm at traffic intersection was designed using Halcon computer vision algorithm library in order to provide a solution for traffic flow detection in intelligent transportation. Application of the moving target detection method based on hybrid Gaussian background model was studied, and the basic algorithm principle was analyzed to select key functions in Halcon for the detection algorithm in HDevelop development environment. In the verification stage, continuous 500 frames of images were processed and analyzed. The result showed that this algorithm could capture moving vehicle targets in the traffic road quite effectively at an average processing speed of 5.8ms per frame, which could meet the real-time requirements.

key words: intelligent transportation; traffic flow detection; machine vision; Halcon