• Acta Optica Sinica
  • Vol. 38, Issue 12, 1228004 (2018)
Feiyan Li*, Hongtao Huo*, Jie Bai, and Wei Wang
Author Affiliations
  • Information Technology and Cyber Security Academy, People's Public Security University of China, Beijing 100038, China
  • show less
    DOI: 10.3788/AOS201838.1228004 Cite this Article Set citation alerts
    Feiyan Li, Hongtao Huo, Jie Bai, Wei Wang. Hyperspectral Target Detection Based on Sparse Representation and Adaptive Model[J]. Acta Optica Sinica, 2018, 38(12): 1228004 Copy Citation Text show less
    (a) Sparse representation without error vector; (b) sparse representation with error vector
    Fig. 1. (a) Sparse representation without error vector; (b) sparse representation with error vector
    Spectral profiles
    Fig. 2. Spectral profiles
    Schematic of similarity calculation
    Fig. 3. Schematic of similarity calculation
    Synthetic data set. (a) Original target for synthesis; (b) synthetic image; (c) target truth map
    Fig. 4. Synthetic data set. (a) Original target for synthesis; (b) synthetic image; (c) target truth map
    (a) AUC value at different sparsity levels; (b) ROC curves at different λ
    Fig. 5. (a) AUC value at different sparsity levels; (b) ROC curves at different λ
    Detection results using different methods. (a) SRTD; (b) SRBBH; (c) SCBD-AWLM; (d) proposed method
    Fig. 6. Detection results using different methods. (a) SRTD; (b) SRBBH; (c) SCBD-AWLM; (d) proposed method
    ROC curves corresponding to different methods
    Fig. 7. ROC curves corresponding to different methods
    (a) AVIRIS dataset and (b) target truth map
    Fig. 8. (a) AVIRIS dataset and (b) target truth map
    (a) AUC value at different sparsity levels; (b) ROC curves at different λ
    Fig. 9. (a) AUC value at different sparsity levels; (b) ROC curves at different λ
    Detection results for AVIRIS dataset. (a) SRTD; (b) SRBBH; (c) SCBD-AWLM;(d) proposed method
    Fig. 10. Detection results for AVIRIS dataset. (a) SRTD; (b) SRBBH; (c) SCBD-AWLM;(d) proposed method
    ROC curves for AVIRIS dataset using different methods
    Fig. 11. ROC curves for AVIRIS dataset using different methods
    MethodAUC
    SyntheticAVIRIS
    SRTD0.56520.7695
    SRBBH0.67520.9327
    SCBD-AWLM0.80780.9119
    Proposed method0.86170.9734
    Table 1. AUC value for different methods and datasets
    Feiyan Li, Hongtao Huo, Jie Bai, Wei Wang. Hyperspectral Target Detection Based on Sparse Representation and Adaptive Model[J]. Acta Optica Sinica, 2018, 38(12): 1228004
    Download Citation