| In order to keep pace with the development of industrial technology and information age,the reform of printing industry is imperative.Color matching as the core step of printing,has also become the main research object.Traditional printing process often relies on manual experience to adjust ink proportion in color matching.Subjective influence and material waste caused by this have become the first step in the reform of printing industry.Due to the improvement of efficiency and quality of intelligent color matching system,it has become the preferred color matching method in various industries.Aiming at the limitation of BP neural network,an ink color matching model based on Stearns-Noechel and BBO optimization algorithm(Biogeography-Based)is proposed.Firstly,the spectral reflectance of color samples and the area ratio of CMYK four-color dots are collected.Stearns-Noechel algorithm mainly finds the functional relationship between different wavelength and spectral reflectance.The function is used as a parameter to modify the spectral reflectance in the algorithm.Then the revised spectral data and CMYK four-color dot area ratio are used as input and output values of BP neural network respectively.Combining the characteristics that organism adapts to living environment in BBO algorithm,the topological structure of BP neural network is optimized(Elite reservation strategy is used in the algorithm to improve the adaptability of the whole environment).Finally,the optimal weights and thresholds are found to achieve the goal of global optimization.There are 6422 sets of experimental data;of which 5600 sets of data are used to train the performance of color matching model and 822 sets of data are used to test.The results show that the average error of BP neural network model without Stearns-Noechel+BBO optimization is 7.9%,while that of the improved model is 3.2%.That is to say,the prediction accuracy of the ink color matching model optimized by Stearns-Noechel+BBO has been greatly improved. |
PDF Full Text Request