• Laser & Optoelectronics Progress
  • Vol. 56, Issue 24, 241004 (2019)
Jianlong Dong1,* and Haoquan Wang2
Author Affiliations
  • 1Academic Affairs Office, Taiyuan Institute of Technology, Taiyuan, Shanxi 0 30008, China
  • 2Key Laboratory of Instrumentation Science and Dynamic Measurement, Ministry of Education, North University of China, Taiyuan, Shanxi 0 30051, China
  • show less
    DOI: 10.3788/LOP56.241004 Cite this Article Set citation alerts
    Jianlong Dong, Haoquan Wang. Elimination of Interferential Text from Images Based on Total Variation Model[J]. Laser & Optoelectronics Progress, 2019, 56(24): 241004 Copy Citation Text show less
    Schematic of area containing texts and its outer neighborhood
    Fig. 1. Schematic of area containing texts and its outer neighborhood
    Schematic of pixels with text interference and their neighborhood points
    Fig. 2. Schematic of pixels with text interference and their neighborhood points
    Image with text interference
    Fig. 3. Image with text interference
    Detection for text interference. (a) Detection result of edge in Fig. 3; (b) result of binarization of Fig. 4(a)
    Fig. 4. Detection for text interference. (a) Detection result of edge in Fig. 3; (b) result of binarization of Fig. 4(a)
    Location of interference text. (a) Connected domain processed by morphology; (b) result after removing connected domain with few pixels
    Fig. 5. Location of interference text. (a) Connected domain processed by morphology; (b) result after removing connected domain with few pixels
    Images obtained by morphological processing and removing small connected domain. (a) Gray scale corresponding to image to be repaired; (b) gray scale after binarization
    Fig. 6. Images obtained by morphological processing and removing small connected domain. (a) Gray scale corresponding to image to be repaired; (b) gray scale after binarization
    Location of interference text. (a) Mask pattern; (b) image with text to be processed
    Fig. 7. Location of interference text. (a) Mask pattern; (b) image with text to be processed
    Result after eliminating interference text in image by using TV model. (a) 5 iterations; (b) 60 iterations
    Fig. 8. Result after eliminating interference text in image by using TV model. (a) 5 iterations; (b) 60 iterations
    Results after eliminating interference text in image by using TV model with different iterations. (a) Image to be processed; (b) 50 iterations; (c) 100 iterations; (d) 300 iterations
    Fig. 9. Results after eliminating interference text in image by using TV model with different iterations. (a) Image to be processed; (b) 50 iterations; (c) 100 iterations; (d) 300 iterations