Abstract | ||
---|---|---|
Text information extracted from scene images is often the key clue for better performance of scene understanding and image retrieval. However, the clutter background and variations, which are intrinsic in scene images, make the natural scene character recognition task rather complicated. To overcome these disadvantages, we propose a novel approach for character recognition task in natural scene images. In the method, character classes are described by groups of local features using a probabilistic model. Structures of characters are represented by mutual positions of local features. For model learning, parameter estimating is done through expectation-maximization in a weak-supervised manner. Experiment results over datasets which includes both synthetic and authentic data demonstrate the validity of the approach. |
Year | DOI | Venue |
---|---|---|
2011 | 10.1007/978-3-642-31919-8_25 | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Keywords | Field | DocType |
scene understanding,novel approach,character recognition task,probabilistic model,local description,character class,natural scene image,scene image,model learning,natural scene character recognition,local feature,expectation maximization,gaussian mixture model | Computer vision,Character recognition,Pattern recognition,Clutter,Computer science,Expectation–maximization algorithm,Image retrieval,Statistical model,Artificial intelligence,Mixture model,Model learning | Conference |
Volume | Issue | ISSN |
7202 LNCS | null | 16113349 |
Citations | PageRank | References |
1 | 0.36 | 7 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
boyu zhang | 1 | 71 | 17.54 |
Wei Zhao | 2 | 47 | 11.11 |
Jiafeng Liu | 3 | 140 | 18.43 |
Rui Wu | 4 | 9 | 5.26 |
Xianglong Tang | 5 | 288 | 44.84 |