Author Affiliations
1 Information Science and Technology College, Dalian Maritime University, Dalian, Liaoning 116026, China2 Collaborative Innovation Research Institute of Autonomous Ship, Dalian Maritime University, Dalian, Liaoning 116026, Chinashow less
Fig. 1. RBD detection model
Fig. 2. Comparison of images before and after Tophat-algorithm preprocessing. (a) Original image; (b) after Tophat-algorithm preprocessing
Fig. 3. Comparison of segmented ships with fixed K value and adaptive K value. (a) Original image; (b) fixed K value; (c) adaptive K value
Fig. 4. Comparison of ship detection from RBD model and improved RBD model. (a) Original image; (b) RBD model; (c) improved RBD model
Fig. 5. Comparison of Otsu segmentation algorithms before and after improvement. (a) Original image; (b) Otsu algorithm; (c) improved Otsu algorithm
Fig. 6. Saliency images without wakes of detections by various algorithms on simple sea surface. (a) Original image; (b) CA algorithm; (c) COV algorithm; (d) GS algorithm; (e) MR algorithm; (f) SF algorithm; (g) RBD algorithm; (h) proposed algorithm
Fig. 7. Saliency images with wakes of detections by various algorithms on sea clutter. (a) Original image; (b) CA algorithm; (c) COV algorithm; (d) GS algorithm; (e) MR algorithm; (f) SF algorithm; (g) RBD algorithm; (h) proposed algorithm
Fig. 8. Saliency images of detections by various algorithms on oily sea surface. (a) Original image; (b) CA algorithm; (c) COV algorithm; (d) GS algorithm; (e) MR algorithm; (f) SF algorithm; (g) RBD algorithm; (h) proposed algorithm
Fig. 9. Saliency images of detections by various algorithms under thin clouds. (a) Original image; (b) CA algorithm; (c) COV algorithm; (d) GS algorithm; (e) MR algorithm; (f) SF algorithm; (g) RBD algorithm; (h) proposed algorithm
Fig. 10. Saliency images of small ship detection by various algorithms. (a) Original image; (b) CA algorithm; (c) COV algorithm; (d) GS algorithm; (e) MR algorithm; (f) SF algorithm; (g) RBD algorithm; (h) proposed algorithm
Fig. 11. Comparison of saliency detection results by different algorithms
Algorithm | Fig. 4 | Fig. 5 | Fig. 6 | Fig. 7 | Fig. 8 |
---|
CA | 37.9 | 40.2 | 40.8 | 39.7 | 41.6 | COV | 23.1 | 25.0 | 25.8 | 24.9 | 26.2 | SF | 11.5 | 12.3 | 12.8 | 11.7 | 13.9 | MR | 6.9 | 7.2 | 7.5 | 7.0 | 7.8 | GS | 6.5 | 7.6 | 7.3 | 6.9 | 7.6 | RBD | 7.8 | 8.3 | 8.5 | 8.0 | 8.8 | Proposed | 11.2 | 12.1 | 12.5 | 11.9 | 13.2 |
|
Table 1. Running time of different modelss