A Method For Color Calibration Based On Simulated Annealing Optimization And Its Application

Achieving a perceptually-consistent color representation is a goal of extreme importance for industrial vision applications.The method for color calibration and other pre-processing technology involved are the necessary prerequisite for subsequent practical application and play important roles.Due to the actual scene is complex,the pre-processing technology must be combined with live thematic studies.The research findings are highly reflected in Nanjing science and technology development projects,such as "Ceramic Tile Sorting System by Visual Sensing",and have important theoretical significance and application value.This paper analyses connotation and position of color calibration method in industrial vision application field.Based on the description and comparison about home and abroad classical color calibration algorithms,we analyzed some deficiencies of the key algorithm.Then we proposed the algorithm and optimization program of related modules includes brightness correction,color calibration based on Simulated Annealing Optimization.The algorithm we proposed achieved better through many specific industrial applications.In terms of the non-uniform about brightness distribution for industrial visual scene,we built a model for correction of brightness space based on the analysis about the Gray Plate easy to fold and the factors that lead the non-uniform about brightness distribution for images.An algorithm procedure about Edge Detection,Morphological Dilation and Saptiotemporal Filter what apply to remove Gray Plate's folds is given in this paper,and the thesis proposes a correction method in brightness space based on CSLIP logarithm computing framework.The proposed method transfers calculation of image pixel values to that of transmittance,then recovers images whose brightness is uniform;In terms of the colour deviation about images for industrial visual scene,this paper analyzes the causes of image acquisition device to produce the phenomenon of color cast and classical color correction algorithm's shortcomings.Then we proposed the method for color calibration based on Simulated Annealing Optimization.By optimizing the idea of the mean value of the color difference,the method is used to improve color accuracy of digital cameras and reconstructed the real color of scene.By experimental verification and practical application,the research results of preprocessing algorithms are credible.The main contributions of this dissertation are summarized as follows:? An algorithm procedure about Edge Detection,Morphological Dilation and Saptiotemporal Filter what apply to remove Gray Plate's folds is improved in this paper,which can solve the problem of the use of the gray plate in the process of folding.? A multi-parameter uneven brightness correction algorithm is improved.CSLIP is applied into brightness correction so that images with even brightness are recovered.? A method for color calibration based on Simulated Annealing Optimization is proposed,that can minimize the mean color errors to achieve higher color fidelity so as to correct the image color to standard white light.? A brightness constraint condition for color calibration is proposed.An automatic luminance constraint strategy is used to make the brightness of the image corrected by the color calibration method remain the same.