Font Size: a A A

Research On Key Technology Of Intraoral Scanning System Based On Binocular Vision

Posted on:2020-04-24Degree:MasterType:Thesis
Country:ChinaCandidate:D S LuFull Text:PDF
GTID:2404330572974638Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of computer technology,digital modulo technology has become the mainstream technology of today.The digital three-dimensional dental mold technology with high precision,easy storage and convenient carrying is also increasingly favored by orthodontists.It not only solves the shortcomings of traditional plaster models,but also improves the patient experience.In this paper,combined with the practical application requirements of clinical stomatology,the key technology research is carried out around the intraoral scanning system based on binocular stereo vision,including camera calibration,polar line correction,binocular stereo matching and tooth 3D model realization.This paper focuses on the above issues.Firstly,the research background of intraoral scanning technology and the research status at home and abroad are introduced.The camera pinhole model,the imaging principle of binocular vision and the plane calibration method based on circular array calibration block are described in detail.Then according to the actual intraoral scanning environment,select the appropriate experimental instrument(binocular camera,calibration block),and finally conduct the calibration experiment of the camera,and obtain the internal and external parameters of the binocular camera,In comparison of the obtained camera internal parameters and external parameters and actual data.The difference between the baseline distance between the two cameras actually obtained and the measured value using the vernier caliper is only 0.912 mm,and the deviation after the polar line correction is 0.422 pixels.the errors are within a reasonable range.Then,the captured image is subjected to distortion and polar line correction by using the internal and external parameters obtained by calibration,and the left and right images are subjected to smoothing filtering and the like.Then the basic knowledge of stereo matching and the traditional local and global stereo matching methods are introduced.The influence of window size on the matching result in local stereo matching algorithm and the matching strategy of this method in real application is studied.The local stereo matching is based on gray.For the special case of no texture,reflection,etc.on the surface of the tooth,it is easy to cause mismatching,and the generated disparity map has a fault and is not smooth enough.In this paper,a variational method based on two-layer W-cycle multi-grid is used for stereo matching between left and right images.The matching cost function is defined by the combination of three different parameters: gray value,gradient value and smoothing term,through the actual stereo matching image pairs to give different weight values,and then through the two-layer W loop multi-grid to optimize the matching cost function to obtain the optimal solution,further improving the continuity of the generated disparity map.The local stereo matching and the algorithm are compared by the square root error and the overall error,and it can be seen that the error rate of parallax graph is about 20% lower than that of traditional local stereo matching in this paper.And for the reconstruction surface with no texture under the narrow baseline,the method is more robust and can obtain a smoother and more accurate disparity map.Finally,the obtained parallax maps are compared by the inverse parallax map to obtain the visual difference map,and the correctness of the disparity map is verified.Finally,the reconstruction method of three-dimensional space points under different binocular stereo vision system is introduced,and the reconstruction of the teeth in the mouth is realized by the parallel configuration of the dual cameras,and the digital three-dimensional dental mold is obtained through the mapping.
Keywords/Search Tags:Dental model, Stereo vision, Binocular calibration, Stereo matching, W-cycle multi-grid
PDF Full Text Request
Related items