2
Instead of conducting borehole scoping underground
with an experienced geologist, one panoramic image can
be generated from a recorded borehole video to visualize
the in-situ structures of borehole walls. Any personnel can
record the borescope video, and a geologist can later review
the borehole after image stitching. The file can be sent to
a geologist remotely, allowing them to provide quick feed-
back even if they are not present at the mine site. There
are mainly two methods to form panoramic borehole
images, namely seamless stitching approach and scan line
approach (Wang et al. 2006). The former method stitches
the video frames with a certain overlapped region to form
a panoramic image with high accuracy, which requires a
large amount of data to be processed. The latter method
uses a designated circle as a scan line to collect the data on
the circle and the collected scan lines are heaped up along
the depth direction to form a borehole image. Due to the
development of image processing techniques and com-
puter sciences, it becomes less difficult to conduct image
stitching, the former method starts to draw attention and
research interests (Niu et al. 2011 Cao et al. 2018 Deng
et al. 2019, 2023 Zou and Song 2021 Zou et al. 2021b, a)
geological structure detection is crucial to the engineering
design and implementation. One of the most commonly
used method is to acquire the borehole videos by Axial
View Panoramic Borehole Televiewer (APBT. Another
advantage of generating panoramic borehole images is
quantitative evaluation. Quantitative measurement can be
conducted for the roof strata within the borehole horizon
after obtaining the panoramic borehole images, and there
are wide applications of borehole images in the literature.
Panoramic borehole images were used to evaluate the rock
mass integrity by avoiding the mechanical disturbance on
cores during the drilling process (Wang et al. 2006). The
structural plane parameters, such as the central position,
orientation and dip angle, and the morphological features
of structural planes in the form of joint roughness coef-
ficient (JRC) can also be extracted from borehole images
(Wang et al. 2017a Zou et al. 2020). The in-situ stress
can be estimated from the recorded change in borehole
shape or borehole breakout in borehole images (Wang et
al. 2018 Zou et al. 2018 Han et al. 2020). The poros-
ity distribution in coral reefs was also evaluated through
high-precision borehole images (Wang et al. 2017b). All
of these available literatures have been focusing on the
forward-view borehole cameras, which captures borehole
images directly ahead of the camera. However, the U.S.
coal industry has been mainly using the side-view bore-
scopes, and there is no available study for the side-view
borescopes.
In a previous study, the authors developed the meth-
odology to generate panoramic borehole images by stitch-
ing the frames of borehole videos recorded with a side-view
borescope, and the video frames recorded a few weeks after
washing were successfully stitched to visualize the borehole
(Xue et al. 2024). However, the image stitching process and
roof lithology identification can be significantly affected by
the borehole wall condition. If a borehole is freshly drilled,
the wall is normally covered with rock dust, making it dif-
ficult to identify the geologic feature and lithology change.
With freshly washed boreholes, water remains on the wall
and reflects light, making it hard to see anything on the
wall. Comparing with these two conditions, the washed
and dried borehole wall was originally considered as the
ideal condition for borehole scoping. In this study, the
influence of borehole wall conditions on the panoramic
borehole images was investigated. The developed meth-
odology for stitching side-view borehole video frames was
first briefly described. The method was then used to gener-
ate panoramic borehole images from the videos recorded
under three different borehole wall conditions. Their influ-
ence was compared based on the success of borehole image
stitching and the identification of geologic features and
lithology change.
METHODOLOGY FOR GENERATING
PANORAMIC BOREHOLE IMAGES
The process for stitching a borehole video is summarized in
the flowchart shown in Figure 1. It normally includes image
acquisition and preprocess, key point detection, key point
matching and filtering, offset calculation, image alignment,
and image blending, if necessary.
Figure 1. Flowchart of image stitching process with a
recorded borehole video
Instead of conducting borehole scoping underground
with an experienced geologist, one panoramic image can
be generated from a recorded borehole video to visualize
the in-situ structures of borehole walls. Any personnel can
record the borescope video, and a geologist can later review
the borehole after image stitching. The file can be sent to
a geologist remotely, allowing them to provide quick feed-
back even if they are not present at the mine site. There
are mainly two methods to form panoramic borehole
images, namely seamless stitching approach and scan line
approach (Wang et al. 2006). The former method stitches
the video frames with a certain overlapped region to form
a panoramic image with high accuracy, which requires a
large amount of data to be processed. The latter method
uses a designated circle as a scan line to collect the data on
the circle and the collected scan lines are heaped up along
the depth direction to form a borehole image. Due to the
development of image processing techniques and com-
puter sciences, it becomes less difficult to conduct image
stitching, the former method starts to draw attention and
research interests (Niu et al. 2011 Cao et al. 2018 Deng
et al. 2019, 2023 Zou and Song 2021 Zou et al. 2021b, a)
geological structure detection is crucial to the engineering
design and implementation. One of the most commonly
used method is to acquire the borehole videos by Axial
View Panoramic Borehole Televiewer (APBT. Another
advantage of generating panoramic borehole images is
quantitative evaluation. Quantitative measurement can be
conducted for the roof strata within the borehole horizon
after obtaining the panoramic borehole images, and there
are wide applications of borehole images in the literature.
Panoramic borehole images were used to evaluate the rock
mass integrity by avoiding the mechanical disturbance on
cores during the drilling process (Wang et al. 2006). The
structural plane parameters, such as the central position,
orientation and dip angle, and the morphological features
of structural planes in the form of joint roughness coef-
ficient (JRC) can also be extracted from borehole images
(Wang et al. 2017a Zou et al. 2020). The in-situ stress
can be estimated from the recorded change in borehole
shape or borehole breakout in borehole images (Wang et
al. 2018 Zou et al. 2018 Han et al. 2020). The poros-
ity distribution in coral reefs was also evaluated through
high-precision borehole images (Wang et al. 2017b). All
of these available literatures have been focusing on the
forward-view borehole cameras, which captures borehole
images directly ahead of the camera. However, the U.S.
coal industry has been mainly using the side-view bore-
scopes, and there is no available study for the side-view
borescopes.
In a previous study, the authors developed the meth-
odology to generate panoramic borehole images by stitch-
ing the frames of borehole videos recorded with a side-view
borescope, and the video frames recorded a few weeks after
washing were successfully stitched to visualize the borehole
(Xue et al. 2024). However, the image stitching process and
roof lithology identification can be significantly affected by
the borehole wall condition. If a borehole is freshly drilled,
the wall is normally covered with rock dust, making it dif-
ficult to identify the geologic feature and lithology change.
With freshly washed boreholes, water remains on the wall
and reflects light, making it hard to see anything on the
wall. Comparing with these two conditions, the washed
and dried borehole wall was originally considered as the
ideal condition for borehole scoping. In this study, the
influence of borehole wall conditions on the panoramic
borehole images was investigated. The developed meth-
odology for stitching side-view borehole video frames was
first briefly described. The method was then used to gener-
ate panoramic borehole images from the videos recorded
under three different borehole wall conditions. Their influ-
ence was compared based on the success of borehole image
stitching and the identification of geologic features and
lithology change.
METHODOLOGY FOR GENERATING
PANORAMIC BOREHOLE IMAGES
The process for stitching a borehole video is summarized in
the flowchart shown in Figure 1. It normally includes image
acquisition and preprocess, key point detection, key point
matching and filtering, offset calculation, image alignment,
and image blending, if necessary.
Figure 1. Flowchart of image stitching process with a
recorded borehole video