中文摘要：

在基于显微磨粒图像处理的磨损故障诊断系统中,图像质量是保证磨粒分析的关键。而磨粒显微图像的成像质量与摄像机参数、光源参数有非常大的关系。采用遗传算法,对摄像机参数和光源参数进行了6个参数自动优化,利用标准图像与某参数下的图像差异构造了适应度函数,通过实数编码、赌轮选择、算术交叉和均匀变异等遗传算法操作,最终得到了一组最优的摄像机和光源参数。在该组参数下,磨粒图像与标准图像达到了最佳的逼近。最后,通过实验验证了该方法的正确有效性。

英文摘要：

Image quality was a key link to ensure the analysis for wear particle in the wear fault diagnosis system which was based on micro-image processing.However,the micro-image quality of wear particle was related to camera parameters and light source parameters.Herein GA was used to adjust 6parameters of camera and light source automatically.A fitness function was constructed to describe the difference between the standard image and the image under certain parameters.A group of optimal parameters of the camera and light source were obtained by means of the GA method operational procedure,such as real coding,roulette wheel selection,arithmetic crossover and uniform mutation operation.Finally,the new method was verified through the experiments,and the results show that the GA method is very effective for adaptive adjustment of multi-parameters in micro-imaging system.