The Application Of Binocular System In Vehicle Information Detection

Binocular system as an extremely important branch of computer vision system,with its convenient operation,good timeliness,high precision,low cost,simple structure,the advantages of the three dimensional range,in the biological medical,aerospace,instrument testing,virtual reality,and other areas of the frontier science and technology,the great and has been widely used.Through the image acquisition system,it obtains several image information of the measured object from different positions,simulates the principle of certain parallax between human eyes when observing the same object,and calculates the difference in position and direction between corresponding points of the image,so as to obtain three-dimensional information of the object.Especially for objects in motion,the acquisition of their image information has higher requirements on the timeliness of the image acquisition system and the calculation and acquisition of three-dimensional information.Therefore,the application of binocular system in vehicle information detection technology has become an effective method.In this paper,according to the characteristics of binocular system,a regular calibrator is selected to calibrate the camera,and the internal and external parameters of the camera are obtained.Then the image was stereo corrected according to the parameters,and the stereo matching of the actual vehicle image was realized by SGBM algorithm,and the parallax map was obtained.Secondly,the depth information between the camera and the vehicle is obtained through the distance measurement principle,and a special marker distance measurement method is used to measure the vehicle's geometric size.At the same time,the vehicle images with obvious representations collected by binocular cameras were characterized and extracted to separate the target area from the background area.Then,based on the improved SURF algorithm,the vehicle characterization initially extracted by the main camera was matched with the original image obtained by the secondary camera to obtain accurate characterization information.In the part of license plate character recognition,the vertical projection method is improved based on license plate features,and an optimized character segmentation algorithm is realized.Finally,taking the license plate as an example,the license plate location result and character recognition result are visualized through programming,thus completing the application of binocular system in vehicle information detection.