Study On Surface Morphology Measurement Technology Based On Structured Light

Unmanned driving technology in intelligent transportation, vehicle safety auxiliary driving, automatic vehicle remote control, factory, patrol and military and other fields have a wide range of applications.At present, the method of three-dimensional detection of obstacles is light flight time, stereo vision, laser line scanning method, structured light scanning method, etc.. Optical flight time, that is, the use of laser radar detection of obstacles. The method can obtain the information of the distance, relative velocity and azimuth angle of the front obstacle.But the size of laser radar is too large, the technology is complex, the equipment is expensive, which limits the popularization of laser radar.Stereo vision method is a common obstacle detection method in the visual navigation of intelligent vehicle. Because of the complexity of the classical image recognition and matching algorithm, the algorithm is difficult to meet the requirements of real-time. Laser line scanning method is a relatively mature three-dimensional acquisition mode, the laser source uses a very thin line laser irradiation to the object, the camera to take pictures, extract the linear laser, get the two-dimensional information, and save the acquired information. For the application of intelligent vehicles, the problem of high accuracy and high cost of the high speed, medium length and long distance, is not easy to be popularized. Structure light scanning method is the projection of the object surface by the projector encoding pattern, with a camera to capture the surface deformation of objects, the use of encoding information to match the image of encoding, and ultimately the use of triangulation principle to calculate the surface space coordinates of objects. It has the advantages of high detection accuracy, but it needs to use a projector to produce a variable structure light, the structure of light intensity can not meet the requirements of detection in the natural environment.In this paper, in order to the demand for unmanned vehicles detected traveling road, a fast road detection technology based on laser dot is proposed.The mathematical model of the pixel coordinates and the spatial coordinates of the pixel coordinates of the projected points on the camera is established, new regional segmentation algorithm is proposed, and the extraction algorithm of the lattice laser center is introduced. The spatial coordinates of the spot are calculated according to the pixel coordinates of the spot, and the three-dimensional inspection reconstruction of the pavement is carried out through the curve fitting of the spot space position,and ultimately the three-dimensional inspection of the pavement.Experimental results show that this method can be realized on the road condition detection, it has a simple structure to detect speed, accuracy and reliability.The method can meet unmanned car driving on the road detection requirements.