Key Techniques For3D Line Structured Light

In recent years, with the rapid development of the machine vision, the visualmeasurement which based on vision measurement technology has become animportant aspect in3D visual measurement. While the line structured lightmeasurement technology has been widely used in industrial production with its imageeasy to extract, high precision, real-time and other characteristics.The line structured light vision measurement technology, which is the laser projectthe beam onto the surface of the object and the camera can get the2D image of thelaser light. we can get the3D information of the object though a series of algorithmsand the model of the line structured light. This paper mainly studies the calibration ofthe measurement system, the line structured light measurement based on the model ofthe glass refraction, the extraction of the laser centerline and unified the globalcoordinates of the multi-vision measurement system.The calibration of the measurement system is mainly divided into the camera’scalibration and the line structure parameters’ calibration of the vision sensor. In thispaper, we use the Zhang’s method to finish the calibration of the cameras. For thereare low accuracy and less calibration points on the laser plane, a method which basedon binocular stereo vision is proposed to solve this problem. we also evaluate theresults.The object is placed on the glass and the laser light enters the camera through theglass will cause refraction when we measure the object. In order to solve this problem,we establish a model of the glass refraction to get the actual3D coordinates of theline structured light. The results show that the proposed method can achieve bettercorrection effect and meet the requirements of engineering applications.The extraction accuracy of the laser centerline affects the accuracy of the finalmeasurement. In this paper, after some image processing to remove noise, we get thesub-pixel coordinates of the centerline by combining with the extreme value methodand the gray-gravity method of the weighted squared. In order to get the full3D information of the object, we build a multi-visionmeasurement system and study the global calibration method based on a commonplane which all the cameras can catch the picture and finish the unified of the globalcoordinates. The experimental results show that the proposed method can quickly andaccurately obtain the3D information of the object, the single measurement RMSerror is less than0.08mm and the standard error is0.199mm to meet the needs of theshoe’s actual production.