Author Affiliations
1Institute of Intelligent Machines, Hefei Institute of Physical Science, Chinese Academy of Science, Hefei 230031, Anhui, China2University of Science and Technology of China, Hefei 230026, Anhui, Chinashow less
Fig. 1. Schematic of weld information measurement
Fig. 2. Projection model measurement (- and are camera coordinate system and image coordinate system, respectively)
Fig. 3. Network structure. (a) Our network; (b)-(d) Detail/Seg Head, ARM, and FFM modules used in the model
Fig. 4. STDC module
Fig. 5. DSNT module
Fig. 6. Training loss of proposed model
Fig. 7. Detail label generation
Fig. 8. Laser stripe segmentation results. (a) Original images; (b) laser stripe label images; (c) weld seam features extracted by FCN-8s; (d) features extracted by our method with detailed information supervision; (e) features extracted by our method without detailed information supervision
Fig. 9. Location results of weld feature points by DSNT method under different noise interferences, where the green and blue “+” are left and right feature points and yellow “+” is intermediate feature point
Fig. 10. Comparison of feature point location results. (a1)-(a3) Extraction errors of left feature point, intermediate feature point, and right feature point in u-axis direction in weld image, respectively; (b1)-(b3) left feature point, intermediate feature point, and right feature point in v-axis direction in weld image, respectively
Fig. 11. Comparison of location errors of subtask correlation feature points. (a) In u-axis direction; (b) in v-axis direction
Fig. 12. Time consumption in processing for image sequence
Stage | Output | Ksize | Stride | Padding | Channel quantity |
---|
ConvX L1 | 60×70 | 3 | 1 | 1 | 64 | ConvX L2 | 60×70 | 3 | 1 | 1 | 32 | Conv2d HM | 60×70 | 1 | 1 | 0 | 3 | DSNT | 3×2 | | | | |
|
Table 1. Feature point positioning branch structure
Model | Resolution /(pixel×pixel) | MIOU /% | FPS /(frame·s-1) |
---|
FCN-8s | 480×560 | 89.07 | 23 | BiSeNetV1 | 480×560 | 96.67 | 42 | BiSeNetV2 | 480×560 | 93.95 | 103 | STDC1-seg | 480×560 | 99.12 | 72 | U-Net | 480×560 | 93.85 | 16 | Ours | 480×560 | 95.97 | 87 | Ours-detail | 480×560 | 98.82 | 87 |
|
Table 2. Comparison of laser stripe segmentation accuracy
Model | Task | Inference time /ms |
---|
Laser fringe segmentation | Feature point location |
---|
FCN-8s | Yes | No | 43.4783 | BiSeNetV1 | Yes | No | 23.6123 | BiSeNetV2 | Yes | No | 9.6583 | ICNet | Yes | No | 48.2594 | Ours | Yes | Yes | 11.4478 |
|
Table 3. Comparison of inference time of different networks