Three-dimensional Texture Profile Measuring System And Contour Reconstruction

3D information of an object reflects the features such as shape contour, surface structure, color texture, etc. This information is helpful for object identification and classification. In recent years, people hope to use electronic devices(such as computer, camera, etc.) and other related technology to complete the information extraction and recognition. Using camera to capture the object images, digital image algorithm and related processing to extract the target 3D information, we can reconstruct the 3D structure of an object, which is called 3D reconstruction based on machine vision.With The modern image processing related hardware function enhancement and cost reduction, reconstruction methods emerge in endlessly, which become the core technology in military reconnaissance, biomedical, and virtual reality(VR). Because of its high lifelike, high automation and high efficiency, it has been widely used in reverse engineering, pattern recognition, industrial inspection, quality control, and other industrial fields. This paper summarizes the applications of the 3D reconstruction technique in industrial measurements and the present domestic and abroad situation and trend of the 3D reconstruction technique, and proposes and implements a new three-dimensional contour and texture measurement system.First, this paper briefly introduces the principle of structure light based on binocular vision for contour detection system, and analyzes the following key technologies: laser lock imaging technology, the camera calibration system, the center line of the line structure light capture. Combining with the best method of band center recognition algorithm, calibrating the camera system is introduced aspect emphatically, and then analysis for the differences of band center related algorithm is presented,And then, we pay much attention to the reconstruction of surface contour and texture details. Three kinds of classical reconstruction algorithm are introduced, and the advantages and disadvantages of the algorithm are summarized. Combined with the actual characteristics of line structured light measurement system principle, a strategy of neighboring contour lines joining together to complete contour surface structure is adopted; In texture reconstruction part we analyze the classic mapping principle, and finally use a special kind of pixel point mapping method for completing this job.Finally, in view of the whole measurement system software realization and reconstruction process, a detailed introduction is presented about the two aspects. In the hardware structure aspect, we point out that the reconstruction system properties, requirements, and clarification of the basic reconstruction process. In another aspect, software, we mainly analyze the data structure(interface), system framework and its overall work process.