Measurement of three-dimensional objects by pattern projection and camera advance
Kai‐Hua Feng, Kōkichi Sugihara, Noboru Sugie · Advanced Robotics · 1989
In studies on vision for intelligent robots one of the most important subjects concerns the extraction from two-dimensional images of three-dimensional (3-D) information such as normals to the object surface and distances from the camera. Numerous methods have been presented so far. A method using cone-shaped beams of light has been proposed by the authors which can measure the normal vectors and the 3-D coordinates of points on the object surface. However, some problems still exist in this method; for example, the non-uniqueness of the solution, measurement errors, resolution of the system, etc. In order to solve these problems, the present study presents a new method which uses both the projection of patterns and the forward movement of the camera. In this method, the changes in the centre positions of the patterns produced by the camera movement are used to determine the depths from the camera. Then the determined depths and apparent distortions of the projected patterns are used for the calculation of the local surface normals. Finally, the solution is improved by using global information of the objects and information on pattern sizes in the central region of the observed scene. Results of experiments with both simulated images and real images are also reported. The present method is advantageous over previous methods in reliability, accuracy, and resolution.