| In recent years, more and more popular in the field of intelligent development, this is theresult of the development of science and technology，this paper puts forward the binocularstereo vision matching on the basis of the two-dimensional image to get three-dimensionalimages of the depth of field information point, more attention paid to the development of thereconstruction of the3D environment, brought new breakthrough for vision research, is animportant research.Feature point matching and area is the heart of a binocular stereo vision matchingtechnology is image matching of this article, also is a difficulty of stereo image matching,should not only consider the feature point matching precision and stability, also want toconsider to the area matching rapidity, effective combination of both, to form a completestereo image matching technology. Feature matching the key algorithm of SIFT featurematching is a novel algorithm, can match the images of the same object for size scale，flipangle change, gray processing and some simple changes in the image processing,the image ofthe information by constraining factor to the perfect match, disruptive feature points out,matching effect is very good. Area matches the image through the gray, denoising, after thematch, sharpening and smoothing pretreatment can greatly improve the speed of matching,the conclusion of the stereo matching for better precision of three-dimensional imageinformation point provides a convenient, played a key role.Regional matching and feature matching combination of matching algorithm not only onthe matching point information fusion the imitation of human visual multi-level classificationsex match factor, the use of polar constraint and disparity gradient constraint operator to pickup the image of the matching point information perfect, more press close to the human eyerecognition technology, but also improves the classification of visual matching rapidprocessing technology, have good robustness and real disparity figure. |
PDF Full Text Request