tailieunhanh - Machine Learning and Robot Perception - Bruno Apolloni et al (Eds) Part 10

Tham khảo tài liệu 'machine learning and robot perception - bruno apolloni et al (eds) part 10', kỹ thuật - công nghệ, cơ khí - chế tạo máy phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | 220 J. Park and G. N. DeSouza by our structured-light scanner. Figure a shows the photograph of the object that was scanned b shows the range image displayed as intensity values c shows the computed 3D coordinates as point cloud d shows the shaded triangular mesh and finally e shows the normal vectors displayed as RGB colors where the X component of the normal vector corresponds to the R component the Y to the G and the Z to the B. Registration Overview A single scan by a structured-light scanner typically provides a range image that covers only part of an object. Therefore multiple scans from different viewpoints are necessary to capture the entire surface of the object. These multiple range images create a well-known problem called registration -aligning all the range images into a common coordinate system. Automatic registration is very difficult since we do not have any prior information about the overall object shape except what is given in each range image and since finding the correspondence between two range images taken from arbitrary viewpoints is non-trivial. The Iterative Closest Point ICP algorithm 8 13 62 made a significant contribution on solving the registration problem. It is an iterative algorithm for registering two data sets. In each iteration it selects the closest points between two data sets as corresponding points and computes a rigid transformation that minimizes the distances between corresponding points. The data set is updated by applying the transformation and the iterations continued until the error between corresponding points falls below a preset threshold. Since the algorithm involves the minimization of mean-square distances it may converge to a local minimum instead of global minimum. This implies that a good initial registration must be given as a starting point otherwise the algorithm may converge to a local minimum that is far from the best solution. Therefore a technique that provides a good initial registration