Intelligent Identification of Coal Crack in CT Images Based on Deep Learning

Abstract

Automatic segmentation of coal crack in CT images is of great significance for the establishment of digital cores. In addition, segmentation in this field remains challenging due to some properties of coal crack CT images: high noise, small targets, unbalanced positive and negative samples, and complex, diverse backgrounds. In this paper, a segmentation method of coal crack CT images is proposed and a dataset of coal crack CT images is established. Based on the semantic segmentation model DeepLabV3+ of deep learning, the OS of the backbone has been modified to 8, and the ASPP module rate has also been modified. A new loss function is defined by combining CE loss and Dice loss. This deep learning method avoids the problem of manually setting thresholds in traditional threshold segmentation and can automatically and intelligently extract cracks. Besides, the proposed model has 0.1%, 1.2%, 2.9%, and 0.5% increase in Acc, mAcc, MioU, and FWIoU compared with other techniques and has 0.1%, 0.8%, 2%, and 0.4% increase compared with the original DeepLabV3+ on the dataset of coal CT images. The obtained results denote that the proposed segmentation method outperforms existing crack detection techniques and have practical application value in safety engineering.

Application of vibration wave CT technology in the dynamic prediction of shock hazards

“…In recent years, as a new geophysical method, seismic wave CT technology has been widely used in engineering and geological diagnosis. A large number of experiments show that CT technology can reconstruct the three-dimensional structure of a coal body when it is damaged (Du et al, 2022) and intelligently identify cracks (Yu et al, 2022). At present, it has become a powerful tool for exploring faults (ZHANG et al, 2020), loose circles (ZHAO and Baojie, 2020), hidden cracks, and stress states in underground mining (Wang et al, 2021).…”

Coal-Rock Crack Image Segmentation Based on Attention Mechanism and Multi-Scale Features

“…Chen et al [15] introduced image processing techniques based on partial differential equations (PDEs) to enable the model to capture foreground edge details while removing noise. Yu et al [16] presented a coal rock crack CT image segmentation framework based on DeepLabv3+ and designed a loss function based on CE Loss and Dice Loss tailored to crack characteristics. Sun et al [17] proposed an attention mechanism fullscale network (FAM-CRFSN) model that extracts coal rock semantic features using a full-scale connection structure and multi-channel residual attention mechanisms.…”