Modified Quadratic Discriminant Functions and the Application to Chinese Character Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
A Survey of Methods and Strategies in Character Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
The Optimum Class-Selective Rejection Rule
IEEE Transactions on Pattern Analysis and Machine Intelligence
Adaptive confidence transform based classifier combination for Chinese character recognition
Pattern Recognition Letters
Precise Candidate Selection for Large Character Set Recognition by Confidence Evaluation
IEEE Transactions on Pattern Analysis and Machine Intelligence
| Hi-index | 0.00 |
In OCR systems the character segmentation algorithmmay generate mis-segmented blocks. Feedbackinformation from character classifier is indispensable toachieve higher character segmentation accuracy. In thispaper a novel rejection algorithm is proposed to identifythese mis-segmented characters more accurately. First,based on confidence evaluation of distance-basedclassifiers, the usual generalized confidence mappingfunction is modified to fit this specific purpose. Second, anovel adaptive thresholding rejection rule is proposed,which is more accurate and flexible. Experiments onChinese, Japanese and Korean document recognitionshowed that new rejection algorithm evidently improvedthe system performance, especially for low-qualityprinted document recognition.