Paper
10 October 2013 Data repair method in incomplete free-surface measurement
Zhongyu Wang, Qiang Li, Hu Yan, Qian Wang
Author Affiliations +
Proceedings Volume 8916, Sixth International Symposium on Precision Mechanical Measurements; 89161V (2013) https://doi.org/10.1117/12.2035595
Event: Sixth International Symposium on Precision Mechanical Measurements, 2013, Guiyang, China
Abstract
Measurement instrument such as coordinate measuring machine or laser scanner is widely applied in the modern industrial manufacturing to reconstruct the shapes of free-surface. However this reconstruction method is determined by the extraction integrity of the shape parameter in the measured part. Some problems occur frequently for the partial surface damage, measurement block, accessible extent, etc., and which result in the measured data incomplete. This paper presents a data repair method based on the gray model combined with neural network theory. The data of the defect surface is divided into different part and the new data sequence is generated. The accumulated generation operation is applied to the new sequence. The normalization processing can then be done before the gray accumulation generation input into the neural network model. The relevant factor sequence by accumulation generation of normalization processing is taken as the RBF neural network input, while the accumulated feature sequence is considered as the output of the network. The defect surface to be repaired is composed of point cloud data. The value of each point can be calculated in the three directions so that the output of the neural network also has three characteristic data. The simulation experimental results show this method can be applied easily in the data repair in the incomplete free-surface with a high accuracy.
© (2013) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Zhongyu Wang, Qiang Li, Hu Yan, and Qian Wang "Data repair method in incomplete free-surface measurement", Proc. SPIE 8916, Sixth International Symposium on Precision Mechanical Measurements, 89161V (10 October 2013); https://doi.org/10.1117/12.2035595
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Neural networks

Data modeling

Clouds

Error analysis

Neurons

Laser scanners

Manufacturing

Back to Top