A Classification Method Based on the Combination of Visible, Near-Infrared and Thermal Infrared Spectrum for Coal and Gangue Distinguishment
SONG Liang1, LIU Shan-jun1*, YU Mo-li1, MAO Ya-chun1, WU Li-xin1, 2
1. College of Resources and Civil Engineering, Northeastern University, Shenyang 110819, China 2. IoT Perception Mine Research Center, China University of Mining and Technology, Xuzhou 221008, China
Abstract：Coals and gangues are the main surface dump in the coal mining process. Dynamic monitoring of those dumps using remote sensing technique is of great importance for mine environmental protection. In the traditional classification of visible and near-infrared remote sensing, part of the gangues might be misclassified as coal, due to the phenomenon of “different objects with the same spectrum”, resulting in the decrease of classification accuracy. Thus, this study firstly acquired visible and near-infrared spectrums of 12 coal samples and 115 gangue samples from Tiefa mining area in China. Most of the gangue samples’ spectrums are different from those of the coals, which can be easily distinguished. While, part of the gangues has the similar spectrum with coal which results in misclassification. With an effort to improve image classification accuracy, furthermore, we acquired the thermal infrared spectrum of the misclassified gangue and the coal samples. The results indicate that there are different spectral characteristics in thermal infrared band between coal and gangue samples, which can be identified easily. Therefore, we proposed a method to separate coal from gangue based on the combination of visible, near-infrared and thermal infrared spectrum. In the first palace, the method conducts measurement on the visible and near-infrared spectrums of all samples for the rough classification recurring to the MAO model. Next, the thermal infrared spectrums of the samples, mixed with gangue and coal are acquired, and the Spectral Absorption Ratio(SAR) is utilized as the evaluation index for the second classification. The fused result of classification originates in the two steps above. The method is further verified by using external samples from Tiefa, Yanzhou, Shendong and Jiangcang mining areas in China, whose results have demonstrated that the method has higher accuracy than that of the traditional classification method based on visible and near-infrared spectrum features. The research results indicates that the conjoint analytical method involving multiple spectrums can solve the phenomenon of “different objects with the same spectrum” in a single band, effectively, which will be of great referential significance in the field of terrain classification based on remote sensing technique.
Key words：Coal;Gangue;Remote sensing;Visible and near-infrared;Thermal infrared
宋 亮1，刘善军1*，虞茉莉1，毛亚纯1，吴立新1, 2 . 基于可见-近红外和热红外光谱联合分析的煤和矸石分类方法研究 [J]. 光谱学与光谱分析, 2017, 37(02): 416-422.
SONG Liang1, LIU Shan-jun1*, YU Mo-li1, MAO Ya-chun1, WU Li-xin1, 2 . A Classification Method Based on the Combination of Visible, Near-Infrared and Thermal Infrared Spectrum for Coal and Gangue Distinguishment. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2017, 37(02): 416-422.
[1] Sun Y Z, Fan J S, Qin P, et al. Environmental Geochemistry and Health, 2009, 1(31): 81. [2] Wang W F, Hao W D, Bian Z F, et al. International Journal of Coal Geology, 2014, 132: 94. [3] Xue Y A, Zhang M M, Li J, et al. Disaster Advances, 2012, 5: 427. [4] Demirel N, Emil M K, Duzgun H S. International Journal of Coal Geology, 2011, 86: 3. [5] Joshi P K, Kumar M, Midha N, et al. Journal of the Indian Society of Remote Sensing, 2006, 4: 415. [6] Schroeter L, Gler C. International Journal of Coal Geology, 2011, 86: 27. [7] WANG Peng, LIU Qing-sheng, LIU Gao-huan, et al(王 鹏, 刘庆生, 刘高焕, 等). Journal of Geo-Information Science(地球信息科学学报), 2013, 15(5): 768. [8] JING Qing-qing, ZHANG Zhi, WANG Xu(荆青青，张 志，王 旭). Coal Science and Technology(煤炭科学技术)，2008, 36(5): 93. [9] Mao Y C, Ma B D , Liu S J, et al. Canadian Journal of Remote Sensing，2014, 40: 327. [10] Smith M R, Bandfield J L, Cloutis E A, et al. Icarus，2013, 223: 633. [11] Sergio S R, María P, Nilda S, et al. Journal of Hydrology，2014, 516: 273. [12] Saljnikov A, Komatina M, Manovic′ V, et al. International Journal of Heat and Mass Transfer，2009, 52: 2871. [13] WANG Jin-nian, ZHENG Lan-fen, TONG Qing-xi(王晋年, 郑兰芬, 童庆禧). Remote Sensing of Environmental, China(环境遥感)，1996, 11(1): 20.
