We present a novel OCR error correction method for languages without word delimiters that have a large character set, such as Japanese and Chinese. It consists of a statistical OC...
—We propose a combinatorial solution to determine the optimal elastic matching of a deformable template to an image. The central idea is to cast the optimal matching of each temp...
—Previous works have shown that the ear is a promising candidate for biometric identification. However, in prior work, the preprocessing of ear images has had manual steps and al...
In this paper model-based reconstruction methods are applied to the detailed reconstruction of buildings from close-range images. The 3D points obtained through image matching are...
This paper presents a language identification technique that differentiates Latin-based languages in degraded and distorted document images. Different from the reported methods tha...