ABSTRACT

Image-based separation of coal and gangue is susceptible to the external environment. This paper studies the effects of illuminance and external moisture on grayscale and texture features of coal and gangue images. We first build an image acquisition system and construct an image dataset of coals and gangues under different illuminance and external moisture. Secondly, we extract four grayscale parameters including gray value corresponding to the maximum frequency, gray average, variance, and skewness and four texture parameters including entropy, contrast, energy, and correlation of these images. Thirdly, we compare these parameters by normalizing the values of coal and gangue. The result shows that illuminance has a greater influence on the grayscale parameters than the texture parameters, and the grayscale maximum fluctuation value is 0.21 while the value of texture is 0.16. External moisture has a greater influence on the texture parameters than the grayscale parameters, and the grayscale maximum fluctuation value is 0.22 while the value of texture is 0.29. The study of the influence of illuminance and external moisture on the features of coal and gangue images could provide an essential guide for image-based identification of coal and gangue under working conditions.

摘要

基于图像的煤矸石分离易受外部环境影响。本文研究了光照和外界水分对煤矸石图像灰度和纹理特征的影响。我们首先构建了一个图像采集系统，并构建了不同照度和外部湿度下煤和矸石的图像数据集。其次，我们提取了四个灰度参数，包括对应于最大频率的灰度值、灰度平均值、方差和偏度以及四个纹理参数，包括这些图像的熵、对比度、能量和相关性。第三，我们通过标准化煤和矸石的值来比较这些参数。结果表明，照度对灰度参数的影响大于纹理参数，灰度最大波动值为0。21 而纹理的值为 0.16。外界水分对纹理参数的影响大于灰度参数，灰度最大波动值为0.22，纹理值为0.29。研究光照和外界水分对煤矸石图像特征的影响，可为工况下煤矸石的图像识别提供必要的指导。