Research On Ranging Algorithm Based On Binocular Stereo Vision Algorithm

Binocular stereo vision technology has been widely used in industrial detection,virtual reality and other fields.In this paper,binocular stereo vision algorithm is studied.The algorithm binocular stereo vision,plane-space algorithm,range scanning algorithm,extreme value elimination method,fly points extreme value elimination method,rotation-plane-space algorithm are proposed.Based on the basic principle of binocular vision algorithm,a two-dimensional spatial similarity transformation model of binocular vision system is set up,and the Zooming Algorithm of binocular stereo vision is obtained by simplifying the two cases of the natural origin reclosing and non coincidence of the sensor.The simulation results show that the calculation results of the shrinkage reduction algorithm are the same as those of the similar transformation model,which proves the rationality of the algorithm.In view of the problem that the Zooming Algorithm does not apply to threedimensional space,a Plane-Space algorithm is proposed.Through the simulation,the simulation data is found to be consistent with the theoretical value,and the rationality of the plane space algorithm is proved.However,when the target distance test system is far away,the experimental data and the simulation data are quite different.Through the analysis of the principle of the image sensor,the reason of the difference between the simulation data and the experimental data of the plane space algorithm is found,and the range calculation method of the binocular stereo vision location algorithm is put forward,and the rationality of the experimental data of the plane space algorithm is proved by an example.The Extreme Value Elimination Method and the Extreme Point Elimination Method are proposed to solve the problem of larger load in the range scanning algorithm.The simulation data of Extreme Value Elimination Method and Flying point Extreme Value Elimination Method show that the boundary of the range boundary and the actual boundary of the plane space algorithm which satisfies the corresponding conditions calculated by the extreme elimination method and the extreme value elimination method of the flying point are consistent with the actual boundary,and the rationality of the extreme elimination method and the extreme value elimination method of the flying point is explained.In view of the problem that the Plane-Space algorithm is not applicable to the cooperative work of multiple binocular stereo vision ranging algorithms under the same standard,a Rotation-Plane-Space algorithm is proposed.The simulation results show that the accuracy of the rotation plane space algorithm is the same as that of the plane space algorithm,which proves the rationality of the Rotation-Plane-Space Algorithm.