基于视觉的车辆衡中轴型检测方法研究

Research on Vision-based Vehicle Shaft Detection Method in Vehicle Scale Application

摘要:

无人化、自动化是现代工业的发展方向,在自动化车辆衡应用场景中,针对现有车辆轴型检测方法中存在设备使用需破坏路面,安装和维护不便的问题,首先根据HOG算子具有较好的对轮胎边缘提取和局部形状信息描述,MB-LBP算子具有抗光照能力强且对小尺度位移具有鲁棒性的特点,设计了一种基于HOG和MB-LBP特征融合的轮胎检测新算法。其次,根据相机成像和图像测距原理,设计了一种基于图像标定和目标检测的轴距测量方法。接着,通过轮胎和轴距的检测结果确定车辆轴型。最后设计了相关实验对提出算法的有效性和准确性进行验证,轴型检测成功率达到97.65%。

Abstract:

Unmanned, automation is the development direction of modern industry, in the application scenarios of automated vehicle scales, aiming at the existing vehicle shaft type detection methods, the use of equipment needs to damage the road surface, and the problems of installation and maintenance are inconvenient, as the HOG operator has better ability to obtain tire edge and describe local shape, and the MB-LBP operator has strong ability to resist light and is robust to small-scale displacement. And a new tire detection algorithm based on HOG and MB-LBP feature fusion was designed for tire detection. Secondly, a wheelbase measurement method based on image calibration and target detection was designed based on the principle of camera imaging and image ranging. Subsequently, determine the axle shape of the vehicle based on the tire and wheelbase test results. Finally, relevant experiments are designed to verify the effectiveness and accuracy performance of the proposed algorithm.The success rate of shaft type detection is 97.65%.