An Improved Descriptor for Chinese Character Recognition

The reference presents a novel approach for Chinese character recognition. Based on it, we know that we can treat character recognition as an image matching problem. Compared with traditional OCR, the new approach for character recognition uniquely uses local invariant descriptors as a new feature extraction method. In this paper, we present a new local descriptor which combines the scale-invariant feature descriptor with contrast distributions of a local region to produce highly efficient feature representation. We extensively evaluated the effectiveness of the new approach with various datasets acquired under varying circumstances. Our experiments demonstrate that our two-component descriptor can represent local region with more information and perform better than SIFT.