1 / 40

A Laplacian Method for Video Text Detection

A Laplacian Method for Video Text Detection. Trung Quy Phan, Palaiahnakote Shivakumara and Chew Lim Tan. Agenda . Introduction Previous Methods Laplacian Method Experimental Results Conclusion and Future Work. Agenda . Introduction Previous Methods Laplacian Method Experimental Results

avalon
Télécharger la présentation

A Laplacian Method for Video Text Detection

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. A Laplacian Methodfor Video Text Detection Trung Quy Phan, Palaiahnakote Shivakumara and Chew Lim Tan

  2. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Conclusion and Future Work

  3. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Conclusion and Future Work

  4. Introduction • Motivation: video indexing • Generates keywords from text • Able to retrieve a particular event or image • Graphic & scene text • Different from camera-based images • Low resolution • Complex background • Text movement & distortion

  5. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Conclusion and Future Work

  6. Previous Methods • Connected component-based • Assumes text pixels have the same colors or grayscale intensities • Edge-based • Works well for high contrast text • Produces false positives for complex backgrounds • Texture-based • Trainable • Computationally expensive

  7. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Conclusion and Future Work

  8. Laplacian Method • Step 1: Text Detection • Identifies candidate text regions • Step 2: Boundary Refinement • Refines the text block boundaries • Step 3: False Positive Elimination

  9. Agenda • Introduction • Previous Methods • Laplacian Method • Text Detection • Boundary Refinement • False Positive Elimination • Experimental Results • Conclusion and Future Work

  10. Text Detection • Text regions have a large number of discontinuities • Input  grayscale  Laplacian-filtered to detect the discontinuities in four directions

  11. Text Detection • Text regions typically have many positive and negative peaks of large magnitudes

  12. Text Detection • Maximum gradient difference (MGD) [1] • For each 1 × N window, MGD is the difference between the maximum and minimum values • Text regions have larger MGD values because of the peaks of large magnitudes

  13. Text Detection • K-means clustering on the MGD map • K = 2, Euclidean distance

  14. Agenda • Introduction • Previous Methods • Laplacian Method • Text Detection • Boundary Refinement • False Positive Elimination • Experimental Results • Conclusion and Future Work

  15. Boundary Refinement • Binary Sobel edge map SM of the input image (only for text regions) • Horizontal and vertical projection profiles

  16. Boundary Refinement • Horizontal • Vertical

  17. Agenda • Introduction • Previous Methods • Laplacian Method • Text Detection • Boundary Refinement • False Positive Elimination • Experimental Results • Conclusion and Future Work

  18. False Positive Elimination • Text block: (1) aspect_ratio ≥ T1 and (2) edge_area / area ≥ T2 • edge_area = number of edge pixels • Otherwise, false positive • T1 = 0.5 and T2 = 0.1

  19. False Positive Elimination • 2 false positives

  20. False Positive Elimination • 1st false positive removed due to the aspect ratio rule

  21. False Positive Elimination • Sobel edge map • 2nd false positive removed due to the edge density rule

  22. False Positive Elimination • Final output

  23. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Conclusion and Future Work

  24. Experimental Results • 101 images: news, sports, movies, etc. • Sizes from 320 × 240 to 816 × 448 • English, Chinese and Korean text • Three implemented methods: edge-based method [1], gradient-based method [2] and uniform-colored method [3]

  25. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Sample Results • Performance Measures • Evaluation • Conclusion and Future Work

  26. Sample Results • Low contrast text Input Edge-based Gradient-based Uniform-colored Proposed

  27. Sample Results • Scene text Input (from [4]) Edge-based Gradient-based Uniform-colored Proposed

  28. Sample Results • The proposed method fails if the contrastis too low Input Proposed Edge-based Gradient-based Uniform-colored

  29. Different font sizes Different languages Sample Results (from [4])

  30. Sample Results • Different window sizes • N = 5 in our experiment Input N = 5 N = 21

  31. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Sample Results • Performance Measures • Evaluation • Conclusion and Future Work

  32. Performance Measures • Detection Rate (DR) • number of localized text / number of text • False Positive Rate (FPR) • number of non-text /number of localized blocks • Misdetection Rate (MDR) • number of text with missing characters / number of localized text DR = 100% FPR = 25% MDR = 67%

  33. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Sample Results • Performance Measures • Evaluation • Conclusion and Future Work

  34. Evaluation

  35. Evaluation • The proposed method outperforms the edge-based and gradient-based methods in all performance measures

  36. Evaluation • Compared to the gradient-based method, the proposed method has a slightly worse MDR but a significantly higher DR

  37. Agenda • Introduction • Previous Methods • Laplacian Method • Experimental Results • Conclusion and Future Work

  38. Conclusion and Future Work • The proposed method performs well on the dataset • Gradient information  candidate text regions • Edge information  localized text blocks • May fail if the contrast is too low • Can be extended for non-horizontal text

  39. References • C. Liu, C. Wang and R. Dai, “Text Detection in Images Based on Unsupervised Classification of Edge-based Features”, ICDAR 2005, pp. 610-614. • E. K. Wong and M. Chen, “A new robust algorithm for video text extraction”, Pattern Recognition 36, 2003, pp. 1397-1406. • V. Y. Mariano and R. Kasturi, “Locating Uniform-Colored Text in Video Frames”, 15th ICPR, Volume 4, 2000, pp 539-542. • X. S. Hua, W. Liu and H. J. Zhang, “Automatic Performance Evaluation for Video Text Detection”, ICDAR, 2001, pp 545-550.

  40. Thank You

More Related