ABSTRACT
A text location algorithm in color images which is suitable to run on entry-level smartphones is proposed to overcome the weaker computing power on smartphones than on computers. In our proposed algorithm, first, morphological approaches are used to do with the edge image and so single text character or text string will form a connected component (CC), namely, the candidate text region. Then, non-text regions are excluded from the candidate text regions by using improved heuristic constraints based on the Gaussian image pyramid and mergence of adjacent connected components which is firstly proposed in this paper. Experiment demonstrates that in terms of the speed, this algorithm runs fast on an entry-level smartphone by only using edge information and in terms of the performance, this algorithm can be used to locate horizontal text lines in different font size, language, and color and it's easy to adapt it to locate vertical text lines. The algorithm effectively solves the problem of high false alarm probability in edge-based methods and gets high precision and recall rate.
- Li, M. H. and Bai, M. 2012. A Novel Text Detection Approach Based on Multi-Structure Element of Morphology. 2012 International Conference on Industrial Control and Electronics Engineering (ICICEE): 1438--1441. DOI= 10.1109/ICICEE.2012.379. Google ScholarDigital Library
- Chen, H., sai, S. T., Schroth, G., Chen, D., Grzeszczuk, R. and Girod, B. 2011. Robust text detection in natural images with edge enhanced maximally stable extremal regions. 2011. Proceedings of International Conference on Image Processing, pp. 2609--2612.Google Scholar
- Yi, C. and Tian, Y. 2011. Text detection in natural scenes by stroke Gabor words. Proceedings of International Conference on Document Analysis and Recognition, pp. 177--181. Google ScholarDigital Library
- Zhou, L., Ping, X. J., Gao, H. L., and Xu, S. 2012. A Novel Video Image Text Detection Method. KSII Transactions on internet and information systems. Volume: 6 Issue: 4 Pages: 1140--1152 DOI:10.3837/tiis.2012.04.011 Published: APR 25 2012.Google Scholar
- Ji, Z., Wang, J., and Su, Y. T. 2009. Text detection in video frames using hybrid features. In Proc. of the 8th International Conference on Machine Learning and Cybernetics, pp.318--342.Google Scholar
- Mancas-Thillou, C., and Gosselin, B. 2006. Spatial and Color Spaces Combination for Natural Scene Text Extraction. Proceedings of IEEE Conference on Image Processing (ICIP), pp. 985--988.Google Scholar
- Yi, C., and Tian, Y. 2011. Text string detection from natural scenes by structure-based partition and grouping. IEEE Trans. Image Process, vol. 20, no. 9, pp. 2594--2605. Google ScholarDigital Library
- Shi, C. Z., Wang, C. H., Xiao, B. H., Zhang, Y., and Gao, S. 2013. Scene text detection using graph model built upon maximally stable extremal regions. Pattern recognition letters. Volume: 34 Issue: 2 Pages: 107--116 DOI: 10.1016/j.patrec.2012.09.019. Google ScholarDigital Library
- Otsu, N. 1979. A threshold selection method from gray-level histograms. IEEE Trans, Syst.Man Cybern, 1979, Vol 9. 62--66.Google Scholar
Index Terms
- Text location in color images suitable for smartphone
Recommendations
Locating text in complex color images
ICDAR '95: Proceedings of the Third International Conference on Document Analysis and Recognition (Volume 1) - Volume 1There is a substantial interest in retrieving images from a large database using the textual information contained in the images. An algorithm which will automatically locate the textual regions in the input image will facilitate this task; the optical ...
Chinese text location under complex background using Gabor filter and SVM
For the Chinese text location under complex background, this paper presents a novel method by combining Gabor filter and support vector machine (SVM). It bases on such a fact that Chinese characters are composed of four kinds of strokes. By extracting ...
A New Approach for Text Location Based on SUSAN and SVM
ISCC-C '13: Proceedings of the 2013 International Conference on Information Science and Cloud Computing CompanionIn order to extract text parts from images, this paper proposes a new approach that combines edge detection, heuristic knowledge and SVM. Firstly, SUSAN algorithm with an adaptive threshold is used to extract the edge information from gray images, and ...
Comments