基于点云处理技术的汽车车身尺寸获取研究

我国处理交通事故的主要方法依然是手工测量,容易产生数据错误与丢失,且效率低下。激光扫描技术是一种以非接触方式快速获取扫描物体表面点云的数据获取技术,而目前我国还没有运用该技术进行交通事故现场数据获取的相关研究。提出一种利用汽车点云获取车身尺寸的方法,并结合车身曲面特点采用适合汽车点云数据预处理的点云处理算法,然后利用matlab软件编写程序获取车身尺寸,最后通过对比实际卷尺测量尺寸与计算所得尺寸,发现两者相差最大值为13 mm,说明该方法可行。

Research on Obtaining the Vehicle Body Size Based on Point Cloud Data Processing Technology

The main method of obtaining data from pedestrian/vehicle collision accidents in our country is manual measurement, which is prone to error and missing data, and it is low efficiency. Laser Scanning Technology is a data acquisition technology which can scan the 3D surface of bodies quickly by non-contact mode, while research has not been done about this technology in traffic accidents. So, a method of obtaining the vehicle body size with vehicle point clouds is put forward, and combined with the characteristics of vehicle surface, an applicable point cloud preprocessing algorithm is adopted, then the vehicle body size is get by using MATLAB. Finally comparing the actual measured sizes and the calculated sizes, it is found that the max error between them is 13mm, so this method is feasible.