基于点云技术的汽车外廓尺寸测量方法研究

Research on Automobile Outline Dimension Measurement Method Based on Point Cloud Technology

针对《机动车安全技术检验项目和方法》(GB 38900—2020)中对于汽车外廓尺寸测量的新要求,采用整车点云数据技术,对牵引车与各类挂车连接后外形上的共性和外廓分割的关键位置进行了研究。通过改进三维点云分割算法,提出一种一并式测量汽车列车和挂车外廓尺寸的测量新方法。该方法可在不改变硬件的前提下,仅通过软件和算法升级,实现一次测量即可测得汽车列车(牵引车、半挂车组合)外廓尺寸参数。以多种结构挂车为试验案例验证了所提方法的可行性,并通过适应性试验分析了所提方法的有效性。研究结果表明,该方法完全能够满足GB 38900—2020中的新增要求,具有广泛的适用性与较高的测量精确度。该方法可将传统方法测量耗时从“小时级”提升为“秒级”,显著提高相关机构的工作效率。

Aiming at the new requirements for the measurement of automobile outline dimension in“Motor Vehicle Safety Technology Inspection Items and Methods”(GB 38900—2020),the commonalities in the shape of tractor-trailers and all kinds of trailers after they are connected,and the key positions of outline segmentation are studied by adopting the whole automobile point cloud data technology.By improving the 3D point cloud segmentation algorithm,a new measurement method is proposed to measure the outline dimension automobile trains and trailers in one and the same way.Without changing the hardware,the method can realize that the parameters of the outline dimension of automobile trains(tractor and semi-trailer combinations)can be measured in one measurement by upgrading the software and algorithms only.The feasibility of the proposed method is verified by using various structural trailers as test cases,and the effectiveness of the proposed method is analyzed through adaptive tests.The research results show that the method can fully meet the new requirements in GB 38900—2020,with wide applicability and high measurement accuracy.The method can improve the traditional methods’measurement time from“hourly”to“secondly”,which significantly improves the working efficiency of related organizations.