• Laser & Optoelectronics Progress
  • Vol. 59, Issue 8, 0810001 (2022)
Qing Yang1、2、*, Li Zhang1, Ran Li2, Bichao Zhan2, Lei Jia2, and Mengyang Liu1、2
Author Affiliations
  • 1Basic Department, Rocket Force University of Engineering, Xi'an , Shaanxi 710025, China
  • 2Beijing Institute of Remote Sensing Equipment, Beijing 100854, China
  • show less
    DOI: 10.3788/LOP202259.0810001 Cite this Article Set citation alerts
    Qing Yang, Li Zhang, Ran Li, Bichao Zhan, Lei Jia, Mengyang Liu. InSAR Terrain Matching Algorithm Based on Morphologically Enhanced HOG Features[J]. Laser & Optoelectronics Progress, 2022, 59(8): 0810001 Copy Citation Text show less
    3D elevation maps. (a) REM; (b) DEM
    Fig. 1. 3D elevation maps. (a) REM; (b) DEM
    HOG feature histograms. (a) REM; (b) DEM
    Fig. 2. HOG feature histograms. (a) REM; (b) DEM
    Diagram of expansion
    Fig. 3. Diagram of expansion
    Diagram of corrosion
    Fig. 4. Diagram of corrosion
    Diagram of open operation effect
    Fig. 5. Diagram of open operation effect
    Diagram of closed operation effect
    Fig. 6. Diagram of closed operation effect
    Two-dimensional elevation maps. (a) REM; (b) DEM; (c) image after closing operation
    Fig. 7. Two-dimensional elevation maps. (a) REM; (b) DEM; (c) image after closing operation
    Line graphs of three kinds of HOG eigenvalues
    Fig. 8. Line graphs of three kinds of HOG eigenvalues
    Flowchart of proposed algorithm
    Fig. 9. Flowchart of proposed algorithm
    Match results. (a) Rough match; (b) smaller match; (c) fine match
    Fig. 10. Match results. (a) Rough match; (b) smaller match; (c) fine match
    Polyline graphs of matching error for different noises. (a) Average matching error in x direction; (b) average matching error in z direction
    Fig. 11. Polyline graphs of matching error for different noises. (a) Average matching error in x direction; (b) average matching error in z direction
    Parameter
    c320.00240.00160.00
    L1 /m720.00840.00960.00
    L2 /m288.00216.00144.00
    L3 /m48.0036.0024.00
    cbin999
    Table 1. Experimental parameter setting
    AlgorithmMatching result of ②
    Error12345Average error
    Gradient cross-correlationΔx /m12.0012.003.006.009.008.40
    Δz /m24.0027.0015.006.006.0015.60
    HOGΔx /m6.003.006.003.006.004.80
    Δz /m15.009.006.006.006.008.40
    EHOGΔx /m0.003.006.003.003.003.00
    Δz /m3.006.009.009.009.007.20
    AlgorithmMatching result of ③
    Error12345Average error
    Gradient cross-correlationΔx /m15.0015.003.009.009.0010.20
    Δz /m6.0015.0027.0030.006.0016.80
    HOGΔx /m6.003.009.003.003.004.80
    Δz /m3.009.0018.0012.006.009.60
    EHOGΔx /m3.003.003.003.003.003.00
    Δz /m3.009.0012.006.006.007.20
    Table 2. Performance comparison of different algorithms
    AlgorithmAverage matching error of noise with different signal-to-noise ratio
    Error9 dB7 dB5 dB3 dB1 dB
    Gradient cross-correlationΔx /m9.6010.2012.6016.8017.40
    Δz /m14.4016.2019.8024.6027.60
    HOGΔx /m4.206.608.4012.6013.20
    Δz /m9.009.6010.2017.4018.00
    EHOGΔx /m2.405.407.2010.2012.00
    Δz /m6.008.409.6014.4016.20
    Table 3. Matching results under different noise conditions
    Qing Yang, Li Zhang, Ran Li, Bichao Zhan, Lei Jia, Mengyang Liu. InSAR Terrain Matching Algorithm Based on Morphologically Enhanced HOG Features[J]. Laser & Optoelectronics Progress, 2022, 59(8): 0810001
    Download Citation