A comparison of statistical regression and neural network methods in modeling measurement errors for computer vision inspection systems

Abstract This paper compares measurement error models for computer vision inspection systems based on the statistical regression method and a neural network-based method. Experimental results demonstrate that both of the models can effectively correct the dimensional measurements of geometric features on a part profile. It also shows that the statistical regression method can perform excellent tasks when the functions for models are carefully selected through statistical testing procedures. On the other hand, varieties of neural network architectures all have good performance when training data are collected carefully. The explicit nonlinear relationship in neural network architectures is very effective in building a general mapping model without specifying the functional forms in advance. While statistical regression methods will continue to play important roles in model building tasks, the neural network-based method will be a very powerful alternative for precision measurement using computer vision systems.

[1]  Roger Mohr,et al.  It can be done without camera calibration , 1991, Pattern Recognit. Lett..

[2]  Chih-Shing Ho.  Precision of Digital Vision Systems , 1983, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  A. W. Burner,et al.  Laboratory calibration and characterization of video cameras , 1990, ISPRS International Conference on Computer Vision and Remote Sensing.

[4]  C. Alec Chang,et al.  Efficient measurement procedures for compound part profile by computer vision , 1991 .

[5]  Faryar Etesami,et al.  Automatic dimensional inspection of machine part cross-sections using Fourier analysis , 1985, Comput. Vis. Graph. Image Process..

[6]  Li Lin,et al.  A neural network-based automated inspection system with an application to surface mount devices , 1993 .

[7]  David Casasent,et al.  Neural net selection of features for defect inspection , 1991, Other Conferences.

[8]  Tariq S. Durrani,et al.  Neural network architecture for circular features extraction in binary patterns , 1991 .

[9]  Oren Masory,et al.  Monitoring machining processes using multi-sensor readings fused by artificial neural network , 1991 .

[10]  Manfred Morari,et al.  PLS/neural networks , 1992 .

[11]  Fred Y. Wu,et al.  Applications of neural network in regression analysis , 1992 .

[12]  Godwin J. Udo,et al.  Neural networks applications in manufacturing processes , 1992 .

[13]  Bernard C. Jiang,et al.  Determine a vision system's 3D coordinate measurement capability using Taguchi methods , 1991 .

[14]  C. Neubauer.  Fast detection and classification of defects on treated metal surfaces using a backpropagation neural network , 1991, [Proceedings] 1991 IEEE International Joint Conference on Neural Networks.

[15]  M. A. Javed,et al.  Neural networks based learning and adaptive control for manufacturing systems , 1991, Proceedings IROS '91:IEEE/RSJ International Workshop on Intelligent Robots and Systems '91.

[16]  Arturo A. Rodriguez,et al.  System calibration and part alignment for inspection of 2-D electronic circuit patterns , 1991, Optics & Photonics.

[17]  Norma Faris Hubele,et al.  Back-propagation pattern recognizers for X¯ control charts: methodology and performance , 1993 .

[18]  Keith L. Peterson,et al.  Counter-propagation neural networks in the modeling and prediction of Kovats indexes for substituted phenols , 1992 .

[19]  Wen-Hsiang Tsai,et al.  Camera Calibration by Vanishing Lines for 3-D Computer Vision , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[20]  James L. McClelland.  Explorations In Parallel Distributed Processing , 1988 .

[21]  Babu Joseph,et al.  Exploratory data analysis: A comparison of statistical methods with artificial neural networks , 1992 .