Abstract
AbstractEnabling machines to read like human beings has been a hot issue for more than fifty years. A novel offline degraded numeral recognition method (DNRBM) based on the measure of medium truth degree (MMTD) is proposed in this paper to identify segmented degraded numeral characters in gray images. It consists of distinguishing foreground from background, rotating an image, wiping off mottles, cutting margins, calculating both statistic and structural features, and recognizing numerals by the fuzzy classifiers constructed based on MMTD using features selected by logistic regression. The experimental results show that in comparison with the template matching method and the k-NN method, the proposed method performs well on recognizing degraded numeral characters with better scalability and better recognition performance.
Highlights
Offline recognition of numerals has many practical applications such as automatic number plate recognition and zip code recognition[1]
We propose a classification approach based on medium truth degree (MMTD) (CBM) to recognize degraded numeral characters (DNRBM)
We have implemented a numeral character recognition system according to the proposed method, and adopt 400 degraded gray images from Ref. 14 as the test dataset, where only a total of 80 relatively legible as well as upright images with 8 samples for each numeral are taken as training samples
Summary
Offline recognition of numerals has many practical applications such as automatic number plate recognition and zip code recognition[1]. Many methods for offline recognition of English, Arabic (Indian), and Persian printed or handwritten numeral characters have been proposed[1,2,3,4,5,6,7,8,9,10]. Only a few of them have taken degraded English numerals into account[4,5,6]. Most methods need a large number of samples for training, which are not always obtainable. This paper aims to address degraded english numeral recognition using proposed fuzzy classifiers fit for a small quantity of training samples, and to test the classification performance of the proposed fuzzy classifiers by taking the numeral recognition as an example. In this paper an integrated method called DNRBM is proposed to identify segmented degraded
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Computational Intelligence Systems
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
