论文信息 - Error compensation for 3D laser scanning system based on neural network

Error compensation for 3D laser scanning system based on neural network

In the field of reverse engineering, high-precision point cloud data is a guarantee of quality and accuracy for three-dimensional (3D) reconstruction model. Aiming to decrease the measuring error of the laser scanning system, an error compensation method for the original point cloud data is proposed. Firstly, the data error is obtained through the comparison between CAD model and original data. Taking the original point cloud data and error data as learning samples, the training work for the BP network is then completed, and the error compensation model is established. Finally, the reliability of the error compensation model is verified by the samples of testing data. With the help of BP compensation model, any original point data error from the laser scanning system can be compensated. And, the results of experiments show the practicality of the method.

Peng Li | Ying Hu | Tianfei Chen

[1] 叶声华,et al. Principle of Coordinates Acquisition Based on Single Camera , 2005 .

[2] Shuzhi Sam Ge,et al. Output feedback adaptive control of a class of nonlinear discrete-time systems with unknown control directions , 2009, Autom..

[3] Martin A. Riedmiller,et al. Advanced supervised learning in multi-layer perceptrons — From backpropagation to adaptive learning algorithms , 1994 .

[4] Guangjun Zhang,et al. A novel calibration approach to structured light 3D vision inspection , 2002 .

[5] Shuzhi Sam Ge,et al. Output Feedback NN Control for Two Classes of Discrete-Time Systems With Unknown Control Directions in a Unified Approach , 2008, IEEE Transactions on Neural Networks.

[6] Ralph R. Martin,et al. Reverse engineering of geometric models - an introduction , 1997, Comput. Aided Des..

[7] Yoo Sang Choo,et al. Leader-follower formation control of underactuated autonomous underwater vehicles , 2010 .