1
24-077
Quantifying the Texture of Coal Images with Different
Lithotypes through Gray-Level Co-Occurrence Matrix
Yuting Xue, Khaled Mohamed, Mark Van Dyke
CDC/NIOSH, Pittsburgh, PA
Dogukan Guner, Taghi Sherizadeh
Missouri University of Science and Technology,
Rolla, MO
ABSTRACT
The Coal Pillar Rib Rating (CPRR) technique has been
developed to assist in rib support design in underground
coal mines. One major challenge of the data collection
process is the measurement of coal strengths in the field.
Schmidt hammer has been verified as a useful tool to deter-
mine coal strength. An alternative approach is to obtain
the representative strength of coal mass by determining
the coal lithotypes in the field based on the coal bright-
ness profile by experienced geologists or mining engineers.
In this paper, image processing techniques have been used
to quantify the texture of coal images of different lithot-
ypes with the purpose of classifying coal lithotypes. The
coal images were collected from the pillar ribs with exposed
surfaces in underground coal mines, and the coal lithot-
ypes were identified when taking the images. The method
of Gray-Level Co-Occurrence Matrix (GLCM) was used to
analyze the textures of coal images of different lithotypes,
and the texture parameters, such as contrast, homogeneity,
energy, and entropy, were compared. The results show that
the images of coal with different lithotypes have different
textures, which can be quantified through the image pro-
cessing. The results from this study demonstrate the poten-
tial of classifying coal lithotypes using rib photos and easing
the data collection process of CPRR.
BACKGROUND
Coal ribs, which are the walls of coal pillars intentionally
retained during mining to support overlying rock strata,
play a crucial role in maintaining the structural integrity
of underground coal mine excavations. The occurrence of
rib falls presents severe safety risks, encompassing injuries,
fatalities, and equipment damage. Researchers from the
National Institute for Occupational Safety and Health
(NIOSH) have been working on the development of an
engineering-based approach for coal rib stability analysis
and support design.
NIOSH researchers have developed the Coal Pillar Rib
Rating (CPRR) technique to quantify the bearing capacity
of coal pillar ribs (Mohamed et al. 2020). The CPRR tech-
nique considers homogeneity, strength, bedding condition,
rock parting condition, face cleat orientation with respect
to entry direction, and coal unit thickness. The calculation
of CPRR requires various input parameters that need to
be measured in the field, like the number of coal units, the
thickness and strength of each coal unit, and the thickness
and strength of rock partings that separate a rib into dif-
ferent coal units. One of the major challenges for the data
collection process of CPRR is the determination of coal
strengths in the field, and Schmidt hammer has been used
as an indirect tool to determine coal strength (Rashed et al.
24-077
Quantifying the Texture of Coal Images with Different
Lithotypes through Gray-Level Co-Occurrence Matrix
Yuting Xue, Khaled Mohamed, Mark Van Dyke
CDC/NIOSH, Pittsburgh, PA
Dogukan Guner, Taghi Sherizadeh
Missouri University of Science and Technology,
Rolla, MO
ABSTRACT
The Coal Pillar Rib Rating (CPRR) technique has been
developed to assist in rib support design in underground
coal mines. One major challenge of the data collection
process is the measurement of coal strengths in the field.
Schmidt hammer has been verified as a useful tool to deter-
mine coal strength. An alternative approach is to obtain
the representative strength of coal mass by determining
the coal lithotypes in the field based on the coal bright-
ness profile by experienced geologists or mining engineers.
In this paper, image processing techniques have been used
to quantify the texture of coal images of different lithot-
ypes with the purpose of classifying coal lithotypes. The
coal images were collected from the pillar ribs with exposed
surfaces in underground coal mines, and the coal lithot-
ypes were identified when taking the images. The method
of Gray-Level Co-Occurrence Matrix (GLCM) was used to
analyze the textures of coal images of different lithotypes,
and the texture parameters, such as contrast, homogeneity,
energy, and entropy, were compared. The results show that
the images of coal with different lithotypes have different
textures, which can be quantified through the image pro-
cessing. The results from this study demonstrate the poten-
tial of classifying coal lithotypes using rib photos and easing
the data collection process of CPRR.
BACKGROUND
Coal ribs, which are the walls of coal pillars intentionally
retained during mining to support overlying rock strata,
play a crucial role in maintaining the structural integrity
of underground coal mine excavations. The occurrence of
rib falls presents severe safety risks, encompassing injuries,
fatalities, and equipment damage. Researchers from the
National Institute for Occupational Safety and Health
(NIOSH) have been working on the development of an
engineering-based approach for coal rib stability analysis
and support design.
NIOSH researchers have developed the Coal Pillar Rib
Rating (CPRR) technique to quantify the bearing capacity
of coal pillar ribs (Mohamed et al. 2020). The CPRR tech-
nique considers homogeneity, strength, bedding condition,
rock parting condition, face cleat orientation with respect
to entry direction, and coal unit thickness. The calculation
of CPRR requires various input parameters that need to
be measured in the field, like the number of coal units, the
thickness and strength of each coal unit, and the thickness
and strength of rock partings that separate a rib into dif-
ferent coal units. One of the major challenges for the data
collection process of CPRR is the determination of coal
strengths in the field, and Schmidt hammer has been used
as an indirect tool to determine coal strength (Rashed et al.