—In this paper, we propose a novel method for extracting handwritten characters from multi-language document images, which may contain various types of characters, e.g. Chinese, ...
Yonghong Song, Guilin Xiao, Yuanlin Zhang, Lei Yan...
A novel text extraction method from graphical document images is presented in this paper. Graphical document images containing text and graphics components are considered as two-d...
We present a method for picture detection in document page images, which can come from scanned or camera images, or rendered from electronic file formats. Our method uses OCR to s...
Functional image classification is the assignment of different image types to separate classes to optimize their rendering for reading or other specific end task, and is an import...
Rafael Dueire Lins, Gabriel Pereira e Silva, Steve...
This paper presents a new background light intensity normalization algorithm suitable for historical document images. The algorithm uses an adaptive linear function to approximate...