CAJ | 학술논문

　　该文提出一种基于计算机视觉的车流量检测方法。首先，利用背景差法对视频中车辆前景图像进行提取。其次，对前景图像进行灰度化、二值化、膨胀、填充等图像处理方法。然后，完成图片中目标块数目的统计，完成车辆的计数。最后，对车辆粘连问题提出了基于车辆形状的解决方案。实验结果表明了该方法的准确性和有效性。
　　해문제출일충기우계산궤시각적차류량검측방법。수선，이용배경차법대시빈중차량전경도상진행제취。기차，대전경도상진행회도화、이치화、팽창、전충등도상처리방법。연후，완성도편중목표괴수목적통계，완성차량적계수。최후，대차량점련문제제출료기우차량형상적해결방안。실험결과표명료해방법적준학성화유효성。
This paper proposes a traffic flow detection method based on computer vision. First, the background subtraction is used to extract vehicles in the foreground images of the video. Second, the methods of grayscale, binarization, expansion and fill-ing are used in the foreground images. Then, complete the counting of vehicles in the images. Finally, the solution based on the shapes of vehicles is proposed to solve the vehicle adhesion problem. The experiment results indicate the accuracy and reliability of the method.