Retrieving information from document images: problems and solutions

Abstract

An information retrieval system that captures both visual and textual contents from paper documents can derive maximal benefits from DAR techniques while demanding little human assistance to achieve its goals. This article discusses technical problems, along with solution methods, and their integration into a wellperforming system. The focus of the discussion is very difficult applications, for example, Chinese and Japanese documents. Solution methods are also highlighted, with the emphasis placed upon some new ideas, including window-based binarization using scale measures, document layout analysis for solving the multiple constraint problem, and full-text searching techniques capable of evading machine recognition errors.

Towards Automatic Blotch Detection for Film Restoration by Comparison of Spatio-Temporal Neighbours

“…Each component is labelled using a contour tracing technique; as polygons are fully determined by vectors, so too are components fully determined by their respective contours, and as such all component pixels can be found. High-order objects such as characters, textlines, and text regions need to be classified in order to effectively perform a task such as DAR [11], but the application of the technique to blotch detection in film sequences appears to be novel.…”

Caption analysis and recognition for building video indexing systems

“…2c). To solve this binarization problem, we propose a binarization method [4,5] that combines a global threshold method with a window-based method. A global threshold T is obtained using Otsu's binarization method [24].…”

“…For this purpose, it is useful to estimate the stroke width of each character and set the window size slightly larger than that width. To estimate the stroke width, one can employ Hadamard multiresolution analysis [4,5]. Hadamard kernels of various scales are applied to each character pixel horizontally and vertically to obtain the strength of each character pixel.…”

“…These works, however, do not have a quantitative basis for evaluating their performance. Our solution to the binarization problem combines the use of the global thresholding method proposed by Otsu [24] and a window-based thresholding method proposed by one of the authors of this paper [4,5]. To evaluate the performance of our hybrid thresholding method, we derive two classifiers from our machine learning method.…”

Real-Time Retrieval for Images of Documents in Various Languages Using a Web Camera

“…For this reason, usage of LLAH for documents in other languages has not been shown yet. If a technique depends on characteristics of a language, applying the technique to different languages sometimes causes difficult problems [2]. However, applying LLAH to other languages is beneficial for users of the languages.…”