基于模拟退火算法优化BP神经网络的色彩空间转换

Optimization of Color Space Conversion of BP Neural Network Based On Simulated Annealing Algorithm

目的研究基于模拟退火算法优化BP神经网络对喷墨打印机色彩空间转换预测准确性的方法。方法通过数据归一化处理、模拟退火算法优化BP神经网络的权值和阈值,以获取它们的全局最优解,再用BP神经网络法进行色差预测。结果模拟退火算法优化BP神经网络预测模型测试15次得到色块平均色差达到2.3067,最小平均色差达到0.7892。结论该方法优化BP神经网络精度非常高,对喷墨打印机色彩空间转换具有较好的非线性拟合能力和更高的预测准确性。

The work aims to study the method of optimizing BP neural network based on simulated annealing algorithm to predict the accuracy of color space conversion in inkjet printer. The weight and threshold of BP neural network were optimized by data normalization and simulated annealing algorithm to obtain the global optimal solution, and then the BP neural network was used to predict the color difference. BP neural network prediction model optimized with simulated annealing algorithm was tested for 15 times to get the average color difference of color lumps. Such color difference reached 2.3067, and the minimum average color difference reached 0.7892. The results show that the proposed method has a high accuracy in optimizing the BP neural network, and has better nonlinear fitting ability and higher prediction accuracy for the color space conversion of inkjet printer.