We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Automatic ISAR Ship Detection Using Triangle-Points Affine Transform Reconstruction Algorithm.
- Authors
Jin, Xinfei; Su, Fulin; Li, Hongxu; Xu, Zihan; Deng, Jie
- Abstract
With the capability of capturing a target's two-dimensional information, Inverse Synthetic Aperture Radar (ISAR) imaging is widely used in Radar Automatic Target Recognition. However, changes in the ship target's attitude can lead to the scatterers' rotation, occlusion, and angle glint, reducing the accuracy of ISAR image recognition. To solve this problem, we proposed a Triangle Preserving level-set-assisted Triangle-Points Affine Transform Reconstruction (TP-TATR) for ISAR ship target recognition. Firstly, three geometric points as initial information were extracted from the preprocessed ISAR images based on the ship features. Combined with these points, the Triangle Preserving level-set (TP) method robustly extracted the fitting triangle of targets depending on the intrinsic structure of the ship target. Based on the extracted triangle, the TP-TATR adjusted all the ship targets from the training and test data to the same attitude, thereby alleviating the attitude sensitivity. Finally, we created templates by averaging the adjusted training data and matched the test data with the templates for recognition. Experiments based on the simulated and measured data indicate that the accuracies of the TP-TATR method are 87.70% and 90.03%, respectively, which are higher than those of the contrast algorithms and have a statistical difference. These demonstrate the effectiveness and robustness of our proposed TP-TATR method.
- Subjects
INVERSE synthetic aperture radar; AUTOMATIC target recognition; SIMULATED annealing; RADAR targets; TRIANGLES; IMAGE recognition (Computer vision)
- Publication
Remote Sensing, 2023, Vol 15, Issue 10, p2507
- ISSN
2072-4292
- Publication type
Article
- DOI
10.3390/rs15102507