Study On The Technology Of Machine Vision In 3D Texture Shape Measurement

Vision allows us more convenient comprehensive understanding the real world. Machine vision imitates the human vision through the electronic means such as computer and camera to perceive and understand the image. 3D texture contour reconstruction is an important issue of the machine vision technology, which has broad applications in culture relics protection, aircraft assembly, reverse engineering and other fields. With the rapid development of vision technology, people are not satisfied with a simple 3D model of object, and begin to pursue the model with the sense of reality. Upto now, 3D reconstruction technique on model reconstruction is often used a pseudo or texture mapping to simulate real surfaces of objects. This technique is unable to provide informations about real texture surface of a 3D object.Based on characteristics of the structured light method, this paper proposes a new texture mapping method to get the 3D real texture information.First of all, this thesis studies the key issues of 3D profile measurement based on the structured light, an image calibration, and an extraction of the center of light stripe. Camera calibration method is mainly studied in detail. Advantages and disadvantages of the commonly algorithm of the light stripe center extraction are analyzed, and a algorithm is selected.Secondly, this thesis studies the 3D shape and texture reconstructions. For the shape reconstruction, based on contours the thesis r introduces a delaunay, BPLI, and a joining algorithms, and analyzes the advantage and disadvantage of related algorithm. Combined the characteristic of structured light measurement, the joining algorithm is adopted to realize the contour curved reconstruction. For the texture reconstruction, the knowledge of texture mapping is represented, and a new method to achieve texture mapping is proposed.Finally, the thesis studies how to realize the whole measurement system from two of direction hardware structure and software design. In the hardware structure, the working process of the measuring is introduced; In the software design, a data structure and the flow chart of each function modul are introduced.