Research On Structured Light Coding And Decoding Technology Of Road Topography Measurement System

The environment perception module is an important component of unmanned driving technology,and unstructured road surface topography measurement is a part of environment perception.Coded structured light technology has lots of benefits,such as low cost,simple structure,and high precision,etc..When applied to the three-dimensional reconstruction of unpredicted roads,it can achieve the goal of unstructured pavement topography measurement.While this method can effectively avoid road traffic risks,it also promotes the application of unmanned driving on unstructured roads.Thus,this paper implemented unstructured pavement topography measurement based on coded structured light measurement technology,and mainly carried out the following work.According to the principle and characteristics of structured light pavement topography measurement,the paper designed the overall technical scheme of the measurement system.Through the comparative analysis of multiple structured light coding schemes,designed the structured light spatial coding strategy based on pseudo-random array combined with binary shape.In the encoding process,in order to obtain a pseudo-random number array template of appropriate size,the folding and splicing algorithm was studied.After the digital array template was obtained,Three kinds of basic symbol structured light coding templates were constructed,which were conducive to the recognition of primitive symbols and the extraction of feature points.The setting of the coding template made the set pixel coordinate points have uniqueness,and solved the matching problem in structured light profile measurement.After projecting the template,the preprocessing and decoding processing algorithms were studied for the structured light image taken by the camera.The preprocessing distinguished the target from the background.In the decoding process,in order to distinguish the primitive symbol pattern,contour detection and improved moment algorithm were studied in turn.On this basis,template matching was achieved,and established the correspondence between the structured light image taken by the camera and the original coding template.The coded structured light measurement system was calibrated based on the perspective projection model,the parameters of the system were obtained.The paper builds coded structured light measurement platform,and system calibration and 3D measurement experiments are carried out.The experiment results show that the hamming distance between sub-windows of coding template is more than 1,which verifies the encoding algorithms are effective,the recognition rate of codeword is over 95%,and the measurement with high accuracy;By experimental verification on actual pavement,the system can satisfy the requirements for measuring the topography of the road surface.