Author Affiliations
1Departanment Information Engineering, Shaoyang University, Shaoyang, Hunan 422000, China2Department of Information Science and Engineering, Hunan First Normal University, Changsha, Hunan 410205, Chinashow less
Fig. 1. Illustration of decomposition of SAR images via BEMD. (a) Original image; (b) BIMF at first level; (c) BIMF at second level; (d) BIMF at third level
Fig. 2. Illustration of classification principle of SVM
Fig. 3. Basic procedure of SAR target recognition method based on BEMD
Fig. 4. Optical and SAR images of ten targets in MSTAR data set. (a) BMP2; (b) BTR70; (c) T72; (d) T62; (e) BRDM2; (f) BTR60; (g) ZSU23/4; (h) D7; (i) ZIL131; (j) 2S1
Fig. 5. Recognition results of proposed method under standard operating condition
Fig. 6. Recognition performances of proposed and compared methods at different pitch angles
Fig. 7. Recognition performances of proposed and compared methods under noise interference
Type | BMP2 | BTR70 | T72 | T62 | BDRM2 | BTR60 | ZSU23/4 | D7 | ZIL131 | 2S1 |
---|
Training | 232(Sn_9566) | 233 | 231(Sn_812) | 299 | 298 | 256 | 299 | 299 | 299 | 299 | Test | 195(Sn_9563) | 196 | 196(Sn_132) | 273 | 274 | 195 | 274 | 274 | 274 | 274 |
|
Table 1. Training and test samples under standard operating condition
Type | Training | Test |
---|
BMP2 | 233(Sn_9563) | 196(Sn_9566)196(Sn_c21) | BTR70 | 233(Sn_c71) | 196(Sn_c71) | T72 | 232(Sn_132) | 195(Sn_812)191(Sn_s7) |
|
Table 2. Training and test samples with configuration variances
Type | Pitchangle /(°) | 2S1 | BDRM2 | ZSU23/4 | T72 |
---|
Training | 15 | 299 | 298 | 299 | 299 | Test | 30 | 288 | 287 | 288 | 288 | | 45 | 303 | 303 | 303 | 303 |
|
Table 3. Training and test samples with pitch angle difference
Method | Average recognition rate /% |
---|
Proposed | 99.12 | SVM-PCA | 97.26 | SRC | 96.36 | CNN | 98.68 |
|
Table 4. Recognition performance of proposed and compared methods under standard opearting condition
Method | Average recognition rate /% |
---|
Proposed | 98.06 | SVM | 95.43 | SRC | 94.88 | CNN | 97.03 |
|
Table 5. Recognition performance of proposed and compared methods under configuration variances