Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis
The separation of coal and gangue is an important process of the coal preparation technology. The conventional way of manual selection and separation of gangue from the raw coal can be replaced by computer vision technology. In the literature, research on image recognition and classification of coal...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2018-03-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | http://www.mdpi.com/2076-3417/8/3/463 |
| _version_ | 1828344036957618176 |
|---|---|
| author | Kai Liu Xi Zhang YangQuan Chen |
| author_facet | Kai Liu Xi Zhang YangQuan Chen |
| author_sort | Kai Liu |
| collection | DOAJ |
| description | The separation of coal and gangue is an important process of the coal preparation technology. The conventional way of manual selection and separation of gangue from the raw coal can be replaced by computer vision technology. In the literature, research on image recognition and classification of coal and gangue is mainly based on the grayscale and texture features of the coal and gangue. However, there are few studies on characteristics of coal and gangue from the perspective of their outline differences. Therefore, the multifractal detrended fluctuation analysis (MFDFA) method is introduced in this paper to extract the geometric features of coal and gangue. Firstly, the outline curves of coal and gangue in polar coordinates are detected and achieved along the centroid, thereby the multifractal characteristics of the series are analyzed and compared. Subsequently, the modified local singular spectrum widths Δ h of the outline curve series are extracted as the characteristic variables of the coal and gangue for pattern recognition. Finally, the extracted geometric features by MFDFA combined with the grayscale and texture features of the images are compared with other methods, indicating that the recognition rate of coal gangue images can be increased by introducing the geometric features. |
| first_indexed | 2024-04-13T23:51:43Z |
| format | Article |
| id | doaj.art-5818474ea9ba4575b2d2dc7af0fa8b15 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-04-13T23:51:43Z |
| publishDate | 2018-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-5818474ea9ba4575b2d2dc7af0fa8b152022-12-22T02:24:04ZengMDPI AGApplied Sciences2076-34172018-03-018346310.3390/app8030463app8030463Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation AnalysisKai Liu0Xi Zhang1YangQuan Chen2School of Mechanical Electronic & Information Engineering, China University of Mining and Technology, Beijing 100083, ChinaSchool of Mechanical Electronic & Information Engineering, China University of Mining and Technology, Beijing 100083, ChinaMechatronics, Embedded Systems and Automation Lab, University of California, Merced, CA 95343, USAThe separation of coal and gangue is an important process of the coal preparation technology. The conventional way of manual selection and separation of gangue from the raw coal can be replaced by computer vision technology. In the literature, research on image recognition and classification of coal and gangue is mainly based on the grayscale and texture features of the coal and gangue. However, there are few studies on characteristics of coal and gangue from the perspective of their outline differences. Therefore, the multifractal detrended fluctuation analysis (MFDFA) method is introduced in this paper to extract the geometric features of coal and gangue. Firstly, the outline curves of coal and gangue in polar coordinates are detected and achieved along the centroid, thereby the multifractal characteristics of the series are analyzed and compared. Subsequently, the modified local singular spectrum widths Δ h of the outline curve series are extracted as the characteristic variables of the coal and gangue for pattern recognition. Finally, the extracted geometric features by MFDFA combined with the grayscale and texture features of the images are compared with other methods, indicating that the recognition rate of coal gangue images can be increased by introducing the geometric features.http://www.mdpi.com/2076-3417/8/3/463coal and ganguefeatures extractionoutline curvefractional calculusmultifractal detrending fluctuation analysis |
| spellingShingle | Kai Liu Xi Zhang YangQuan Chen Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis Applied Sciences coal and gangue features extraction outline curve fractional calculus multifractal detrending fluctuation analysis |
| title | Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis |
| title_full | Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis |
| title_fullStr | Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis |
| title_full_unstemmed | Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis |
| title_short | Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis |
| title_sort | extraction of coal and gangue geometric features with multifractal detrending fluctuation analysis |
| topic | coal and gangue features extraction outline curve fractional calculus multifractal detrending fluctuation analysis |
| url | http://www.mdpi.com/2076-3417/8/3/463 |
| work_keys_str_mv | AT kailiu extractionofcoalandganguegeometricfeatureswithmultifractaldetrendingfluctuationanalysis AT xizhang extractionofcoalandganguegeometricfeatureswithmultifractaldetrendingfluctuationanalysis AT yangquanchen extractionofcoalandganguegeometricfeatureswithmultifractaldetrendingfluctuationanalysis |