Related Experiment Video
Updated: May 23, 2025
13:19
Deep Neural Networks for Image-Based Dietary Assessment
Published on: March 13, 2021
8.9K
A large-scale open image dataset for deep learning-enabled intelligent sorting and analyzing of raw coal
Ziqi Lv1,2, Yuhan Fan3,4, Te Sha5
1School of Chemical & Environmental Engineering, China University of Mining & Technology (Beijing), Beijing, 100083, P. R. China. lvziqi@cumtb.edu.cn.
Scientific Data
|March 8, 2025
Summary
This study introduces DsCGF, a large-scale raw coal image dataset crucial for advancing intelligent coal sorting in China. The dataset supports deep learning models for accurate coal, gangue, and foreign object identification.
Area of Science:
- Energy Science
- Computer Vision
- Materials Science
Background:
- China's energy transition emphasizes intelligent sorting of raw coal for carbon goals.
- Deep learning advancements are pivotal for intelligent coal preparation.
- A lack of accurate, large-scale data hinders intelligent coal sorting progress in China.
Purpose of the Study:
- To introduce DsCGF, a large-scale, open-source raw coal image dataset.
- To address the data scarcity issue in intelligent coal preparation.
- To facilitate research in intelligent raw coal sorting using computer vision.
Main Methods:
- Systematic collection and meticulous annotation of raw coal images over five years.
- Dataset creation from three representative mining regions in China.
- Annotation at multiple levels for coal, gangue, and foreign objects, supporting classification, detection, and segmentation tasks.
Main Results:
- Development of DsCGF, a dataset with over 270,000 visible-light images.
- Multi-level annotations for coal, gangue, and foreign objects.
- Dataset designed for image classification, object detection, and instance segmentation.
Conclusions:
- The DsCGF dataset effectively supports research in intelligent raw coal sorting.
- This resource is vital for advancing deep learning applications in coal preparation.
- DsCGF contributes to China's energy transition goals through improved coal sorting accuracy.
