Low Resolution Character Recognition by Image Quality Evaluation

The character image database plays an important role for the evaluation of a character recognition system. But there is no measure which tells the level of recognition difficulty of a given database. This paper proposes a novel approach for the low resolution character recognition, which fits the input character for the appropriate character database according to the input image quality. It is composed of two stems: character image quality evaluation, character recognition. Firstly, it presents the gray distribution feature to evaluate the character image quality. Secondly, according to the evaluation result the appropriate character database and recognition method are selected for the input character image which makes the classification have the higher probability of being the correct decision. Experiment results demonstrate the proposed approach highly improved the performance of the degraded character recognition system