Title :
Handwriting Character Recognition as a Service: A New Handwriting Recognition System Based on Cloud Computing
Author :
Gao, Yan ; Jin, Lanwen ; He, Cong ; Zhou, Guibin
Author_Institution :
Sch. of Electron. & Inf. Eng., South China Univ. of Technol., Guangzhou, China
Abstract :
In this paper, we propose a new framework of providing Handwritten Character Recognition as a Service (HCRaaS) via Internet, based on cloud computing technology. Using the proposed Cloud-based recognition platform, we would be able to apply many advanced algorithms in practice, such as modified quadratic discriminant function (MQDF) and SVM classifier used for large scale character recognition, writing adaptation technology, and handwriting Chinese word/text line recognition that usually involve large storage and computation cost. With the merits and characteristics of HCRaaS, users of different mobile devices are not only no longer subject to local computing capacity and storage resource constraints, but also, they can benefit from higher recognition accuracy and personalized service with low hardware costs. The experimental results show that, the proposed HCRaaS system based on Cloud computing can provide reliable handwriting solution across different mobile OS with higher recognition performance.
Keywords :
cloud computing; handwritten character recognition; image classification; mobile computing; natural language processing; support vector machines; text analysis; HCRaaS system; SVM classifier; cloud based recognition platform; cloud computing; computation cost; computing capacity; handwriting Chinese word recognition; handwritten character recognition as a service; large scale character recognition; mobile OS; mobile devices; modified quadratic discriminant function; storage resource constraints; writing adaptation technology; Character recognition; Cloud computing; Feature extraction; Handwriting recognition; Mobile communication; Servers; Writing; Cloud computing; HCRaaS; handwriting recognition;
Conference_Titel :
Document Analysis and Recognition (ICDAR), 2011 International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4577-1350-7
Electronic_ISBN :
1520-5363
DOI :
10.1109/ICDAR.2011.181