不同照度下煤矸图像灰度及纹理特征提取的实验研究

研究煤和矸石在不同照度下的响应特性，设计了差异照度煤矸图像采集系统，进行了煤矸图像采集实验，建立了煤矸灰度子图像数据库，对数据库中的子图像的灰度、纹理特征进行了提取，通过定义归一化特征差异指数以及基于支持向量机(SVM)对实验数据进行了分析与讨论。结果表明:在同一照度下，煤和矸石在灰度及纹理特征上存在差异，且随着照度的改变，煤和矸石的特征也发生变化，同时两者的变化规律存在显著差异。基于不同特征的SVM分类器的识别正确率不同，照度的变化也会对分类器的识别正确率产生影响;当考虑照度因素后，分类器的识别正确率最大增加了13.31%，此外基于多特征多照度融合的SVM分类器性能较好，识别正确率为98-39%。

Experimentalstudyongrayandtexturefeaturesextractionofcoalandgangueimageunderdifferentilluminance

In order to study the response characteristics of coal and gangue under different illuminance,a coal gangue image acquisition system with adjustable illuminance was designed,and some coal gangue image acquisition experiments were carried out.A coal gangue gray sub image database was established,and the gray and texture features of the sub images in the database were extracted.By defining the Normalized Feature D value Index (NFDI) and training the support vector machine (SVM) classifier,the experimental data were analyzed.The results show that coal and gangue exhibit some different gray and texture characteristics under the same illuminance,and these characteristics change with the change of illuminance.The recognition accuracies of SVM classifiers based on different characteristics are different.When the illuminance factor is taken into consideration,the recognition accuracy of SVM based on gray and texture features is improved.In addition,the SVM classifier based on the multiple features and illuminance fusion shows a good performance,and the recognition accuracy reaches 98-39%.