Author Affiliations
1State Key Laboratory of Modern Optical Instrumentation, College of Optical Science and Engineering, Zhejiang University, Hangzhou, Zhejiang 310027, China2Laser Fusion Research Center, China Academy of Engineering Physics, Mianyang, Sichuan 621900, Chinashow less
Fig. 1. Imaging principle of laser scanning confocal microscope. (a) Device principle; (b) pinhole structure; (c) horizontal and vertical scanning
Fig. 2. Two modes of silica detected by laser scanning confocal microscope. (a) Scattering mode; (b) fluorescence mode
Fig. 3. Typical subsurface defects collected by confocal microscope. (a) Pitting defect; (b) pit defect
Fig. 4. Processing effects of Fig. 3(b) by using defect enhancement algorithm. (a) Enhanced image; (b) aggregated image
Fig. 5. Principle of double-threshold aggregation algorithm. (a) Original image; (b) processed image
Fig. 6. Principle of improved MC algorithm based on octree algorithm. (a) Establishment of volume data; (b) octree segmentation
Fig. 7. Flow chart of three-dimensional reconstruction algorithm
Fig. 8. Reconstruction of pit defect in simulation. (a) Simulated defect; (b) reconstructed defect in simulation; (c) residual of reconstruction
Fig. 9. Tomography image obtained from simulation. (a) Cross-section of defect from simulation; (b) confocal tomography image obtained from simulation
Fig. 10. Restoration ratio of point cloud after reconstruction by three different algorithms
Fig. 11. Detection results of subsurface defects. (a) Scratch defect; (b) microcrack defect; (c) pit defect
Fig. 12. Subaperture scanning images
Fig. 13. Reconstruction results of subsurface defects. (a)(b) Reconstruction results of scratch defects; (c)(d) reconstruction results of microcrack defects; (e)(f) reconstruction results of pit defects
Fig. 14. Destructive test results of subsurface defects of fused silica. (a) Etching test results of pit defects
[17]; (b) polishing-residual subsurface defects
[18]; (c) scratch defects
[19]; (d) pit defects
[20] Number of voxels | Average time /ms | Memory space consumption /MB | |
---|
Contour Filter | | Original MC | Improved MC | Contour Filter | Original MC | Improved MC |
---|
103 | 0.5 | 0.4 | 0.4 | 0.01 | 0.01 | 0.01 | 503 | 70 | 51 | 45 | 2.1 | 3.1 | 1.3 | 1003 | 312 | 356 | 286 | 101 | 140 | 32 | 10002×100 | 19312 | 28632 | 10832 | 752 | 900 | 562 |
|
Table 1. Running time and occupied memory spaces of three different algorithms
Defect size /μm3 | Number of defects | Total volume /μm3 | Volume ratio (0--50 μm in depth) /% |
---|
0--100 | 13 | 556 | 0.002 | >100--200 | 7 | 1123 | 0.003 | >200--300 | 21 | 5433 | 0.012 | >300--400 | 24 | 8512 | 0.025 | >400 | 6 | 3362 | 0.010 |
|
Table 2. Defect volume distributions of experimental samples