Acta Optica Sinica, Volume. 29, Issue 2, 412(2009)
Phase-Height Mapping Algorithm Based on Neural Network
Establishing high precision phase-height mapping is one of the key techniques in structural light measurement system. Based on establishing accurate camera image and projector image correspondence, the three-layer back propagation neural network is trained to build a mapping relationship between image coordinates and three-dimensional coordinates. A plane block with circle marks is used to collect sample data and train the neural network. In order to verify the precision of this algorithm, a standard sphere and a plaster model are measured using the trained network. The experimental results show that the algorithm proposed in this work can measure complex free-form surface objects. The measurement precision can achieve 0.095 mm.
Get Citation
Copy Citation Text
Li Zhongwei, Wang Congjun, Qin Dahui, Shi Yusheng. Phase-Height Mapping Algorithm Based on Neural Network[J]. Acta Optica Sinica, 2009, 29(2): 412