论文部分内容阅读
为了提高智能交通系统中运动车辆检测的准确率,提出了一种基于张量恢复的APG-TR算法。采用张量表征交通视频图像,保持视频图像高维结构特征。通过张量恢复,重建出张量的低秩部分与稀疏部分,实现交通视频图像中交通背景与运动目标车辆的分离与交通视频内在特征的提取。利用交通监控系统采集到的交通视频106帧图像对本文算法进行了测试。测试结果表明:在晴天条件下,APG-TR算法的平均正确率为91.4%,在雨、雾天气条件下,正确率分别为86.4%、85.2%,相比帧差法更加稳定与准确。APG-TR算法具有良好的收敛速度与鲁棒性,在智能交通领域中具有广泛的应用前景。
In order to improve the accuracy of moving vehicle detection in intelligent transportation system, an APG-TR algorithm based on tensor restoration is proposed. Tensor is used to characterize traffic video images and maintain the high dimensional structure of video images. Through the tensor recovery, the low-rank part and the sparse part of the tensor are reconstructed to separate the traffic background and the moving target vehicle in the traffic video image and to extract the intrinsic features of the traffic video. Traffic images collected by traffic monitoring system are used to test the algorithm in this paper. The test results show that the average correct rate of APG-TR algorithm is 91.4% in sunny days and 86.4% and 85.2% respectively in rainy and foggy weather conditions, which is more stable and accurate than the frame difference method. APG-TR algorithm has good convergence speed and robustness, and has a wide range of applications in the field of intelligent transportation.