• Laser & Optoelectronics Progress
  • Vol. 56, Issue 1, 011005 (2019)
Shanshan Du and Chao Han*
Author Affiliations
  • College of Electrical Engineering, Anhui Polytechnic University, Wuhu, Anhui 241000, China
  • show less
    DOI: 10.3788/LOP56.011005 Cite this Article Set citation alerts
    Shanshan Du, Chao Han. Total Variation Inpainting Algorithm Based on Logarithmic Operation and Additional Directional Gradient Operator[J]. Laser & Optoelectronics Progress, 2019, 56(1): 011005 Copy Citation Text show less
    wP curves of TV algorithm
    Fig. 1. wP curves of TV algorithm
    Curve of PSNR
    Fig. 2. Curve of PSNR
    Curve of time
    Fig. 3. Curve of time
    wP curves of proposed algorithm
    Fig. 4. wP curves of proposed algorithm
    Curve of PSNR
    Fig. 5. Curve of PSNR
    Curve of time
    Fig. 6. Curve of time
    First simulation results. (a) Text image; (b) mask image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Fig. 7. First simulation results. (a) Text image; (b) mask image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Second simulation results. (a) Damaged image; (b) TV algorithm; (c) algorithm in Ref. [12]; (d) algorithm in Ref. [18]; (e) proposed method
    Fig. 8. Second simulation results. (a) Damaged image; (b) TV algorithm; (c) algorithm in Ref. [12]; (d) algorithm in Ref. [18]; (e) proposed method
    Third simulation results. (a) Fissured image; (b) mask image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Fig. 9. Third simulation results. (a) Fissured image; (b) mask image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Fourth simulation results. (a) Boat; (b) damaged image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Fig. 10. Fourth simulation results. (a) Boat; (b) damaged image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Fifth simulation results. (a) Man; (b) damaged image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Fig. 11. Fifth simulation results. (a) Man; (b) damaged image; (c) TV algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [18]; (f) proposed method
    Sixth simulation results. (a) Texture image; (b) damaged image; (c) n=2000; (d) n=3000; (e) n=5000
    Fig. 12. Sixth simulation results. (a) Texture image; (b) damaged image; (c) n=2000; (d) n=3000; (e) n=5000
    ImageParameterTV algorithmAlgorithm in Ref. [12]Algorithm in Ref. [18]Proposed method
    Fig. 7n5000500020001000
    t /s45.2500411.968822.437512.8906
    Fig. 8n30010010030
    t /s1.500025.34380.92192.3594
    Fig. 9n200020001000200
    t /s51.8906900.756354.9375365.5156
    Table 1. Comparison of iteration times and time for four kinds of inpainting algorithms
    ImageParameterTV algorithmAlgorithm in Ref. [12]Algorithm in Ref. [18]Proposed method
    Fig. 10PSNR /dB31.144430.470430.909531.2535
    n100103510
    t /s4.38132.10941.79690.7269
    Fig. 11PSNR /dB42.475638.866642.165242.9750
    n80030010030
    t /s31.546951.71883.93751.5313
    Table 2. Comparison of PSNR, iteration times and time for four kinds of inpainting algorithms
    Shanshan Du, Chao Han. Total Variation Inpainting Algorithm Based on Logarithmic Operation and Additional Directional Gradient Operator[J]. Laser & Optoelectronics Progress, 2019, 56(1): 011005
    Download Citation