Feature-specific illumination patterns for automated visual inspection

The choice of an appropriate illumination design is one of the most important steps in creating successful machine vision systems for automated inspection tasks. In a popular technique, multiple inspection images are captured under angular-varying illumination directions over the hemisphere, which yields a set of images referred to as illumination series. However, most existing approaches are restricted in that they use rather simple and generic illumination patterns on the hemisphere. Furthermore, the spectrum of the illumination is assumed to be fixed and is not considered. In this paper, we present an illumination technique which reduces the effort of capturing a series of inspection images for individual reflectance features by using linear combinations of basis light patterns that vary in their directional and spectral radiance. The key idea is to encode linear functions for feature extraction as angular- and wavelength-dependent illumination patterns, and thereby to compute linear features from the scene's spectral reflectance field directly in the optical domain. Finally, we evaluate and verify the proposed illumination technique to the problem of optical material type classification of printed circuit boards (PCBs).

[1]  Hans-Peter Seidel,et al.  Realistic, hardware-accelerated shading and lighting , 1999, SIGGRAPH.

[2]  Tim Weyrich,et al.  Principles of Appearance Acquisition and Representation , 2009, Found. Trends Comput. Graph. Vis..

[3]  Stephen H. Westin,et al.  Predicting reflectance functions from complex surfaces , 1992, SIGGRAPH.

[4]  Steven D. Brown Introduction to Multivariate Statistical Analysis in Chemometrics , 2010 .

[5]  Bernd Jähne,et al.  Learning of Optimal Illumination for Material Classification , 2010, DAGM-Symposium.

[6]  Kostas Daniilidis,et al.  Catadioptric Projective Geometry , 2001, International Journal of Computer Vision.

[7]  Vincent Masselus,et al.  A Practical Framework for Fixed Viewpoint Image-Based Relighting (Praktische modellen voor beeldgebaseerde herbelichting met vast camerastandpunt) , 2004 .

[8]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[9]  Robin Gruna,et al.  Acquisition and evaluation of illumination series for unsupervised defect detection , 2011, 2011 IEEE International Instrumentation and Measurement Technology Conference.

[10]  Paul E. Debevec,et al.  Acquiring the reflectance field of a human face , 2000, SIGGRAPH.

[11]  Oliver Wang,et al.  Material classification using BRDF slices , 2009, CVPR.

[12]  Fernando Puente León,et al.  Model-Based Segmentation of Surfaces Using Illumination Series , 2007, IEEE Transactions on Instrumentation and Measurement.

[13]  Matthew O'Toole,et al.  A Basis Illumination Approach to BRDF Measurement , 2010, International Journal of Computer Vision.