openpits asbestos

Published: 3 February 2023| Version 2 | DOI: 10.17632/pfdbfpfygh.2
Mikhail Ronkin,


Database includes images of open-pit places taken in the Bazhenovskoye field, Russia. All images are taken in the different weather and day conditions. All data are labeled for instance segmentation (as well as object detection) problems and have labeling in the COCO format. The archive contains both: all data in the images folder and annotation in the annotations folder. The labeling was performed manually in the CVAT software. The image size is 2592 × 2048.


Steps to reproduce

Data Set was collected from 220 images taken in the open pit during 10 experiments in the Spring, Summer, Autumn and Winter time and the different weather (sunny, cloudy, snow). Each experiment was containing 3 - 5 open pit processing places chosen as most representable by geological service of the field processing company. For each open-pit place several images of its different parts were taken. Each image contains about 20-80 fragments of different sizes. For each of the collected data the instance segmentation labeling was carried out using the CVAT software. Most of the presented images was taken using the following system: electrically adjustable turntable platform maintained on the tripod; camera for computer vision "Dalsa Genie Nano M2590NIR'' with the gray-scale matrix 1 inch and resolution 2590 x 2048 pixels (5MP), and enhanced sensitivity in the near-infrared range; lens "LMZ25300M3P-IR'' with electrically adjustable 12-times zoom and enhanced sensitivity in the near-infrared range; infrared backlight with wavelength 850 nm and with the manually adjustable zoom set within the distance 5--10 m with lighting angle of 30 degrees; supply battery for autonomous work and PoE system (Power over Ethernet) for camera supply. The selected camera and lens allow obtaining a resolution of about 4 pixels in 1 mm at a distance of 5 m that is assumed to be enough in comparison with the typical asbestos vein width (about 4--12 mm).


Ural'skij federal'nyj universitet imeni pervogo Prezidenta Rossii B N El'cina


Computer Vision, Object Detection, Coal Fragmentation, Mining, Deep Learning, Instance Segmentation


This research was supported by the Russian Science Foundation and Government of Sverdlovsk region, Joint Grant No 22-21-20051,