Research on Intelligent Detection and Segmentation of Rock Joints Based on Deep Learning

Table 1

The mAP (mean average precision) of different algorithms.

Algorithm	mAP of detection box (%)	mAP of segmentation (%)
The algorithm proposed in this paper	58.0	49.2
Mask R-CNN	47.3	38.1
Cascade R-CNN	49.5	38.8
Yolact	45.5	35.2
Mask Scoring R-CNN	50.2	41.1

Bold values represent the model evaluation metrics obtained using the algorithm proposed in this paper.