Research On Depth Measurement Of Laser Vision Indoor Scene Based On Embedded System

With the development of computer vision technology,it is of great significance to obtain the depth information of indoor 3d scenes from 2d images in the fields of target detection,mobile robot positioning and navigation,and indoor 3d reconstruction.This thesis designs an indoor scene depth measurement system based on embedded platform and line laser vision measurement technology.This system first uses a line laser and a camera to scan the indoor scene in three dimension,collects video data containing the line laser fringe,then extracts the sub-pixel coordinates of the line laser stripe center through image preprocessing and the light bar center positioning algorithm,and finally uses the triangulation principle calculates the depth information of each point on the surface of the object in the scene.The main contents of this thesis are as follows: Firstly,this thesis studies the realization principle of laser vision system and the mathematical model of laser triangulation ranging,which is introduced from the point laser ranging model to the line laser ranging model;Secondly,based on the laser vision measurement theory,the four preprocessing algorithms of image adaptive cropping,graying,median filtering and Otsu threshold binarization are studied to eliminate image noise interference,realize the rough extraction of line laser light bars,and propose improved Gray center of gravity method to extract sub-pixel center of light bar.Then,the system's software and hardware measurement platform is built.The hardware platform is mainly composed of USB camera,line laser,servo and embedded ARM platform.The software platform is developed based on embedded Linux system and Open CV,including scanning control program and image acquisition program And image processing program;Next,on the basis of building the hardware and software platform,the measurement system correction was performed,including camera image distortion correction and key hardware parameter correction.In this thesis,the camera plane calibration method was used to complete the distortion correction,and the hardware parameters were corrected by curve fitting.At the end of this thesis,the measurement system was tested,designing a scheme to collect depth information at different angles and different distances.The experimental results verify the feasibility of the depth measurement system designed in this thesis.