Computer Science and Information Engineering

Fen Cai1, Yan-Min Luo2, Pei-Zhong Liu3 and Hsuan-Ming Feng This email address is being protected from spambots. You need JavaScript enabled to view it.4

1College of Mathematics and Computer Science, Quanzhou Normal University, Quanzhou, Fujian 362000, P.R. China
2College of Computer Science & Technology, Huaqiao University, Xiamen 361021, P.R. China
3College of Engineering, Huaqiao University, Quanzhou, Fujian 362021, P.R. China
4Department of Computer Science and Information Engineering, National Quemoy University, Kinmen, Taiwan, R.O.C.

 

Received: November 30, 2017
Accepted: September 11, 2018
Publication Date: December 1, 2018

Download Citation: ||https://doi.org/10.6180/jase.201812_21(4).0012  

ABSTRACT

Detecting and locating text information in complex background images has become an important technology for marking the image automatically. The traditional text segmentation method is difficult to obtain the ideal result in the text segmentation of complex background images. According to the features of complex background images, this paper puts forward a new segmentation method by fusing the edge and region features. Firstly, the initial localization of text region is extracted by using stroke filter based on the characteristics of Chinese characters. Secondly, the method of projection histogram analysis is used to remove the conglutination between Chinese characters and background, and the connected region analysis method based on region growing is used to recognize the single or multiple active targets. Finally, the final text segmentation is performed by the improved vertical projection segmentation algorithm which based on the structure of Chinese. The experimental results show that the proposed algorithm has good performance in text location, and can adapt to the characteristics of the complicated background and get accurate character image area comparing with the traditional edge detection method.


Keywords: Edge Feature, Region Feature, Stroke Filter, Text Segmentation


REFERENCES

  1. [1] Wang, Y., H. Zheng, and D. W. Hu (2004) The Technology to Acquire Text in Image and Vedio, Journal of Image and Graphic 9(12), 532538.
  2. [2] Zhong, Y., K. Karu, and A. K. Jain (1995) Locating Text in Complex Color Images, Proceedings of the 3rd International Conference on Document Analysis and Recognition 28(10), 15231535. doi: 10.1109/ ICDAR.1995.598963
  3. [3] Botz-Bornstein, T., and M. M. Mostafa (2017) A Corpus-based Computational Analysis of Philosophical Texts: Comparing Analytic and Continental Philosophy,International Journal of Social and Humanistic Computing 2(3), 230246. doi: 10.1504/IJSHC.2017. 10005737 [4] Zhu, C. J., C. Li, and Z. Xiong (2007) Text Detection and Recognition Technique in Video, Computer Engineering 33(10), 218219.
  4. [5] Keechul, J., K. Kwang, and K. J. Anil (2004) Text Information Extraction in Images and Video: a Survey, Pattern Recognition 37(5), 977997. doi: 10.1016/j. patcog.2003.10.012
  5. [6] Lienhart, R., and A. Wernicke (2002) Localizing and Segmenting Text in Images and Videos, IEEE Transactions on Circuits and Systems for Video Technology 12(4), 256268. doi: 10.1109/76.999203
  6. [7] Mariano, V. Y., and R. Kasturi (2000) Locating Uniform-colored TextinVideo Frames,IEEEProceedings of the 15th International Conference on Pattern Recognition 539542. doi: 10.1109/ICPR.2000.902976
  7. [8] Zhong, Y., H. J. Zhang., and A. K. Jain (2000) Automatic Caption Localization in Compressed Video, IEEE Transactions on Pattern Analysis and Machine Intelligence 22(4), 385392. doi: 10.1109/34.845381
  8. [9] Pan, Y. F., X. Hou, and C. L. Liu (2011) AHybrid Approach to Detect and Localize Texts in Natural Scene Images, IEEE Transactions on Image Processing 20(3), 800 813. doi: 10.1109/TIP.2010.2070803
  9. [10] Zhang, X., and F. C. Sun (2011) Pulse Coupled Neural Network Edge-based Algorithm for Image Text Locating, Tsinghua Science & Technology 1(16), 2230. doi: 10.1016/S1007-0214(11)70004-9
  10. [11] Liu, P. Z., X. Ning, and W. J. Li (2014) Approach for Text Segmentation in Web Image, Computer Engineering and Applications 50(10), 180184.
  11. [12] Shi, Z. G., and L. Q. Gao (2010) New Algorithm for Text Segmentation Based on Stroke Filter, Computer Science 37(1), 287289.
  12. [13] Liu, C. P., G. Liu, C. H. Wang, and Y. Q. Zhu (2013) Design of ImageText Extracting ExperimentBased on Stroke Filter, Research and Exploration in Laboratory 11, 8992.
  13. [14] Liu, Q., C. Jung, and Y. Moon (2006) Text Segmentation Based on Stroke Filter, Proceedings of ACM International Conference on Multimedia, USA, 129 132.
  14. [15] Liu, Q., C. Jung, S. Kim, et al. (2006) Stroke Filter for Text Localization in Video Images, Image Processing, 2006 IEEE International Conference on. IEEE, 14731476. doi: 10.1109/ICIP.2006.312560
  15. [16] Yao, J. L., L. B. Weng, and X. H. Wang (2012) A Text Region Location Method Based on Connected Component, Pattern Recognition and ArtificialIntelligence 25(2), 325331.
  16. [17] Cao, W. (2010) Character Vertical Projection Segmentation Based on Dichotomy, Software Cuide 10, 7172.
  17. [18] Shi, J., and J. Malik (2000) Normalized Cuts and Image Segmentation, IEEE Transactions on Patten Analysis Machine Intelligence 22(8), 888905. doi: 10. 1109/34.868688

Latest Articles

Multi-level Semantic Fusion Network for Few-shot Multimedia Image Recognition in Education Management
Multi-level Semantic Fusion Network for Few-shot Multimedia Image Recognition in Education ManagementVolume 28, Issue 2 (In Press)

Read more: ...

River bank erosion prediction using multivariable linear regression
River bank erosion prediction using multivariable linear regressionVolume 27, Issue 12 (In Press)

Read more: ...

Optimization of Ultrasound-Assisted Pectin Extraction from Durian Rind
Optimization of Ultrasound-Assisted Pectin Extraction from Durian RindVolume 28, Issue 2 (In Press)

Read more: ...

Short-time prediction model for cross-section passenger flow in urban transit trains using GCN-AM-BiLSTM
Short-time prediction model for cross-section passenger flow in urban transit trains using GCN-AM-BiLSTMVolume 28, Issue 2 (In Press)

Read more: ...

Response Prediction Analysis of RC Frame Structures Using Support Vector Machine Algorithm
Response Prediction Analysis of RC Frame Structures Using Support Vector Machine AlgorithmVolume 28, Issue 2 (In Press)

Read more: ...

Using the non-dominated sorting genetic algorithm II for multi-objective optimization of acoustic performance in sandwich panels with cellular honeycomb core
Using the non-dominated sorting genetic algorithm II for multi-objective optimization of acoustic performance in sandwich panels with cellular honeycomb coreVolume 28, Issue 2 (In Press)

Read more: ...

Analyzing Glass Configurations For Energy Efficiency In Building Envelopes: A Comparative
Analyzing Glass Configurations For Energy Efficiency In Building Envelopes: A ComparativeVolume 28, Issue 2 (In Press)

Read more: ...

Assessment of the optimal gold recovery rate through heap leach extraction process from hydrothermal alteration zone ore
Assessment of the optimal gold recovery rate through heap leach extraction process from hydrothermal alteration zone oreVolume 28, Issue 2 (In Press)

Read more: ...

Research on the Time-Varying Relationship between Macroeconomic Variables and the Stock Market Volatility Based on TVP-VAR Model
Research on the Time-Varying Relationship between Macroeconomic Variables and the Stock Market Volatility Based on TVP-VAR ModelVolume 28, Issue 2 (In Press)

Read more: ...