Reaserch Of Stereo Matching Technology Based On Open CV And Coded Structured Light

Stereo matching is one of the most important aspects of binocular vision, thepurpose of which is to find corresponding relationship of pixels in images. Stereomatching contributes to calculate the disparity map and providing three-dimensionaldata for reconstruction. Stereo matching algorithm consists of region matching, featurematching, phase matching dynamic programming, graph cuts and so on. As there aremany uncertainties in taking camera image, a universal algorithm is not exist, which canmeet various requirements of stereo matching, now the popular stereo algorithm run insome certain conditions. Any kind of stereo matching algorithm is to settle two basicrequirements, that is to improve matching accuracy and induce processing speed.Gray Code is a time encoding technology, applied widely in structured3Dmeasurement. In the paper we proposed a stereo matching algorithm based on GrayCode, which based on blending the Gray Code3D measurement and stereo matchingtechniques. The method can reduce the difficulty of feature, improve the accuracy offeature extraction and ensure the accuracy of stereo matching, by increasing its surfacecharacteristics through projecting Gray Code on the object which be measured. At thesame time, the epipolar constraint reduces the search range from the whole image to alimited number of points which have same code value. It can reduces the research timein searching match point and improve the speed of stereo matching.The proposed algorithm is achieved on software platform, which is established byvisual C++6.0and Open CV stereo visual type library. Open CV is an open codesource of visual library developed by Intel, it provides some general image processingand compute vision algorithm, by the form of functions to users. It can greatly simplifythe programming complexity, increase development speed and shorten the developmentcycle. The experiment in the paper verified the feasibility of the algorithm, which get abetter matching result, the accuracy and speed of matching had improved.