Chenran Li
ABSTRACT
In order to improve the efficiency of image content retrieval, combined with computer vision technology, an intelligent key information detection and recognition system applied to images is proposed. The system is divided into four parts: positioning the text area, detecting key information, recognizing key information and broadcasting voice. Firstly, the image is preprocessed to obtain a clear binary image. Secondly, the dynamic line segmentation algorithm is used to obtain the candidate text areas of the image, and the non-text areas are filtered out with the revelatory rule. Then, the Bayesian classification algorithm is used to discriminate the candidate text region to obtain the key information. Finally, the image of key information is converted into editable text through Optical Character Recognition technology, and the text is converted into speech using text-to-speech conversion technology. The experimental results show that the recognition system of key information is simple and efficient, and the classification accuracy is good, which is suitable for image key information recognition.
- Zhong Y, Karu K, Jain A K, “Locating text in complex color images,” in Proceedings of the IEEE International Conference on Document Analysis and Recognition, Quebec, Montreal ,1995, vol. 1, pp. 146- 149.Google Scholar
- Cai Bo, Zhou Dongru, Hu Hongbin, “Research and implementation of subkey information detection and extraction in digital video,”J. Journal of Computer-Aided Design & Computer Graphics, vol. 15, no. 7, pp. 898-903, 2003.Google Scholar
- Li M, Wang C, “An adaptive text detection approach in images and video frames,”in Proceedings of the IEEE International Joint Conference on Neural Networks,Hong Kong, China,2008, vol. 1, pp. 72-77.Google Scholar
- Epshtein B, Ofek E, Wexler Y, “Detecting text in natural scenes with stroke width transform,”in Proceedings of the IEEE Computer Vision and Pattern Recognition, San Francisco, CA, USA, 2010, vol. 1, pp. 2963-2970.Google Scholar
- Neumann L, Matas J, “ A Method for Text Localization and Recognition in Real-World Images,” in Proceedings of the IEEE Computer Vision, Queenstown, New Zealand, 2010, vol. 1, pp. 770- 783.Google Scholar
- Liu Shuqin, “ Image feature extraction method and its application,” Ph.D. dissertation, Dept. Elect. Eng, Northwest Univ, Xi'an, Shaanxi ProvinceGoogle Scholar
- Vladimir Kozoderov,Timofei Kondranin,Egor Dmitriev,Vladimir Kamentsev, “Bayesian classifier applications of airborne hyperspectral imagery processing for forested areas,”J. Advances in Space Research,vol.55, no.11, pp. 2657-2667, 2015.Google ScholarCross Ref
- Wang Chunfeng, Zhang Yonghong, “Bayesian network structure learning method based on unconstrained optimization and genetic algorithm,” J .Control and Decision, vol. 28, no. 4, pp. 618-622, 2013.Google Scholar
- Zhang Xinyu, “Bayesian Maximum Likelihood Estimation for Combination Forecasting,” J .Forecasting, no.5, pp. 48-68, 1998.Google Scholar
- Larrañaga P, Karshenas H, Bielza C, , “ A review on evolutionary algorithms in Bayesian network learning and inference tasks,”J. Information Sciences, vol. 233, no. 2, pp. 109-125, 2013.Google ScholarDigital Library
- Jiang Hongyi, Zhu Liping, Ou Zhangpeng, “ Text recognition of natural scene images based on MSER and Tesseract,” J. Computer Knowledge and Technology, vol. 13, no. 33, pp. 213-216, 2017.Google Scholar
- Bi Xiaojun, Jing Guangyu, Xu Xianfeng, “Speech synthesis of text files using TTS technology,” J. Automation Technology and Application, vol. 23, no. 9, pp. 49-51, 2004.Google Scholar
Recommendations
MAPS: midline analysis and propagation of segmentation
ICVGIP '12: Proceedings of the Eighth Indian Conference on Computer Vision, Graphics and Image ProcessingScenic word images undergo degradations due to motion blur, uneven illumination, shadows and defocussing, which lead to difficulty in segmentation. As a result, the recognition results reported on the scenic word image datasets of ICDAR have been low. ...
A Bayesian-based method of unconstrained handwritten offline Chinese text line recognition
This paper presents a new Bayesian-based method of unconstrained handwritten offline Chinese text line recognition. In this method, a sample of a real character or non-character in realistic handwritten text lines is jointly recognized by a traditional ...
A knowledge-based recognition system for historical Mongolian documents
This paper proposes a knowledge-based system to recognize historical Mongolian documents in which the words exhibit remarkable variation and character overlapping. According to the characteristics of Mongolian word formation, the system combines a ...
Comments