采用图像识别技术对LC4CS铝合金周期降雨模式加速腐蚀实验试样的腐蚀形貌进行了研究,提出了图像特征值δ的概念.以特征值δ作为考察周期降雨实验总体加速性的指标,分析了正交实验各实验因子对加速性的影响,并得到加速腐蚀实验的最佳水平组合,其结论与根据腐蚀失重数据分析所得的结论一致.
Image recognition was used to study the acceleratively corroded morphology of LC4CS aluminum alloy samples with periodic rain tests, and a parameter, δ, describing image feature was introduced. The effects of orthogonal experiment factors on acceleration were analyzed by parameter 5, and the best combination of factor lever for accelerated corrosion test was also obtained. The conclusion is consistent with the result based on the corrosion weight loss.