Three-dimensional shape measurement using binary spatio-temporal encoded illumination

A new approach for three-dimensional shape measurement is proposed. The corresponding point pairs between the projector and camera are identified by projecting a special pseudorandom sequence onto the surface of measured objects. The important properties of this sequence are expressed as follows: (1) any subsequence with a length of four symbols is unique; (2) there are no repeated symbols in any subsequence. The pseudorandom sequence is constructed with an alphabet of six symbols. These symbols are encoded with local spatial and temporal information by pixels in vertical strips of the projector plane. During measurement, the patterns for encoding are projected in turn and are captured with a camera. The projected pseudorandom sequence is retrieved by analysing the captured images. Then, the corresponding point pairs are worked out. Finally, the shapes of objects are reconstructed with triangulation. The experimental results reveal that a dense depth image with high resolution can be obtained. However, texture can be acquired simultaneously. The proposed method is robust.