Color mixing and random search for optimal illumination in machine vision
Authors:
- 1 December 2013
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 907-912
Abstract
This study proposed how to find optimal illumination for industrial vision in short time using random search algorithm and multiple color light sources. The fineness of an image captured by a monochrome camera is varied by illumination and can be evaluated by image sharpness. The relation between the sharpness and the illumination is non-linear, so direct optimum methods are applicable to mix the multiple sources. Random search is one of the direct optimum methods and were derived from the sharpness as input and N driving voltages for N light sources. The random search was tested in an RGB mixer and reduced the number of iteration for optimal illumination compared with conventional equal step search.Keywords
This publication has 19 references indexed in Scilit:
- Mixed-color Illumination and Quick Optimum Search for Machine VisionInternational Journal of Optomechatronics, 2013
- An Optical Mixer and RGB Control for Fine Images using Grey Scale DistributionInternational Journal of Optomechatronics, 2012
- From active contours to active surfaces2011
- Multi-Spectral Imaging by Optimized Wide Band IlluminationInternational Journal of Computer Vision, 2008
- Three-color mixing for classifying agricultural products for safety and qualityApplied Optics, 2006
- Application of color mixing for safety and quality inspection of agricultural products2005
- Autofocusing in computer microscopy: Selecting the optimal focus algorithmMicroscopy Research and Technique, 2004
- Probabilistic detection and tracking at high frame rates using affine warping2003
- Colored Illumination for Enhancing Discriminability in Machine VisionJournal of Visual Communication and Image Representation, 1995
- Geometric primitive extraction using a genetic algorithmIEEE Transactions on Pattern Analysis and Machine Intelligence, 1994