Abstract

Text extraction from scene images is a defy subject in light of low resolution, complex background and textual style/text size varieties. In this paper, we design a scheme to detect text based on shape features like Euler Number, a number of pixels for each region which candidate to be a character and vertical distances as a geometric feature between these regions. We divide these features into base features to collect the text regions, and the other features as a filter to discard the non-text regions. We use some threshold with the features either to extract text regions or to discard non-text regions. The proposed method outperforms some existed method through the basis metric.

References

• Qixiang Ye and David Doermann, Text Detection and Recognition in Imagery: A Survey, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 37, no. 7, pp. 1480-1500, June 2015.
• Chunmei Liu and et al., Text Detection in Images Based on Unsupervised Classification of Edge-based Features, in Proceedings of Eight International Conference on Document Analysis and Recognition, 2005, pp. 610-614
• Shivakumara P. and et al., Accurate video text detection through classification of low and high contrast images, Pattern Recognition, 43(2010), 2165-2185.
• Jie Y. and et al., A method for text line detection in natural images , Multimed Tools Appl (2015) 74: 859-884.
• Kita K. and Toru W., Binarization of Color Characters in scene images Using k-means Clustering and Support Vector Machines, International Conference on Pattern Recognition, 2010.
• Honggang Zhang and et al, Text extraction from natural scene image: A survey, Neurocomputing, 122,(2013)
• Shivakumara P. and et al., Multioriented Video Scene Text Detection Through Bayesian Classification and Boundary Growing, IEEE Transaction on Circuits and systems for Video Technology, vol. 22, no. 8, 2012.
• Shivakumara P. and et al, A new multi-modal approach to bib number/text detection and recognition in Marathon images, Pattern Recognition, 61, pp. 479-491, 2017.
• Sue We and Adnan Amin, Automatic Thresholding of Gray-level Using Multi-stage Approach, Proceedings of the Seventh International Conference on Document Analysi Recognition, IEEE, 2003.
• Abdel-Rahiem Hashem and et al., A Comparison study on text detection in scene images based on connected component analysis, IJCSIS, vol. 15, no. 2, pp. 127-139, 2017
• Matias Valdenegro- Toro and et al, Histogram of Stroke Width for Multi-script Text Detection and Verification in Road Scenes, IFAC , 2017.
• Shi C. and et al., End-to-end scene text recognition using tree-structure models, Pattern Recognition, 47, pp. 2853-2866, 2014
• Yu Qiao and et al., Thresholding based on variance and intensity contrast, Pattern Recognition, 40, pp. 596-608, 2007
• Regionprops. Measure properties of image regions https://www.mathworks.com/help/images/ref/regionprops.html
• Quartile. From Wikipedia: https://en.wikipedia.org/wiki/Quartile.