沥青混合料油石比的图像识别方法

Image recognition method of asphalt-aggregate ratio for asphalt mixture

为了快速方便地检验沥青混合料油石比,选取普通沥青混合料AC-13C,按不同油石比击实成型标准马歇尔试件,并对试件进行水平切割,采用数码相机拍摄沥青混合料切面的数字照片,据此通过图像处理提取沥青灰度特征。分析结果表明:沥青混合料切面图像的灰度频率分布曲线具有典型的双峰分布特征,通过采用加权的双正态分布概率密度函数拟合左峰半枝灰度频率分布曲线,所提取的沥青灰度特征明显,相应得到的识别沥青用量与实际油石比之间具有显著的线性相关性,判定系数超过0.99;图像识别方法的识别油石比与实际油石比之间的绝对误差一般在±0.1%以内。可见,采用图像处理技术识别沥青混合料油石比可行。

In order to verify the asphalt-aggregate ratio of asphalt mixture in a rapid and convenient way, the common asphalt mixture AC-13C was selected to shape standard Marshall specimens at different asphalt-aggregate ratios with compaction method. The specimens were cut horizontally, the digital photos of cross sections for asphalt mixture were photographed with a digital camera, and asphalt gray features were extracted by using image processing method. Analysis result shows that there is a typical double-peak distribution feature for the gray frequency distribution curve in the cross section image of asphalt mixture. The weighted probability density function of double-normal distribution is applied to fit the left peak semi-branch curve of gray frequency distribution, and the extracted asphalt gray features are obvious. There is a significant linear correlation between the obtained recognition asphalt contents and the practical asphalt-aggregate ratios, and the determination coefficient is more than 0.99. The absolute errors between the recognition asphalt-aggregate ratios of image recognition method and the practical asphalt-aggregate ratios are generally within !0. 1 ~. So the application of image processing technology is feasible to recognize the asphalt-aggregate ratio of asphalt mixture. 3 tabs, 9 figs, 18 refs.