Fig. 1. Computation of index function
Fig. 2. Weight computation of feature points
Fig. 3. Homomorphic shrinkage algorithm. (a) Before simplification; (b) after simplification
Fig. 4. Schematic of retention index
Fig. 5. Flow chart of the proposed algorithm
Fig. 6. Raw data. (a) kerolamp model; (b) helix model
Fig. 7. Delaunay triangulation construction. (a) kerolamp model; (b) helix model
Fig. 8. Piecewise linear falling Morse complex extracted by the proposed algorithm. (a) kerolamp model; (b) helix model
Fig. 9. Piecewise linear falling Morse complex extracted by the method in Ref.[
24]. (a) kerolamp model; (b) helix model
Fig. 10. Feature extraction results for kerolamp model in different thresholds. (a) λ=0.8,XT =0.23; (b) λ=0.8,XT =0.3; (c) λ=0.8,XT =0.32
Fig. 11. Feature extraction results for helix model in different thresholds. (a) λ=0.7,XT=0.15; (b) λ=0.7,XT=0.2; (c) λ=0.7,XT=0.22
Fig. 12. Feature extraction results for model 1 in different noise levels. (a) 0 dB; (b) 10 dB; (c) 20 dB; (d) 50 dB
Method | Number of extracted features | Runing time /s | Point cloud compression ratio /% |
---|
Proposed method | 1068 | 0.8 | 28.75 | Method in Ref.[24] | 1405 | 2.7 | 6.27 |
|
Table 1. Comparison of different methods for kerolamp model