We describe a multi-perspective vision studio as a flexible high performance framework for solving complex image processing and machine vision problems on multi-view image sequences. The studio abstracts multi-view image data from image sequence acquisition facilities, stores and catalogs sequences in a high performance distributed database, allows customization of back-end processing services, and can serve custom client applications, thus helping make multi-view video sequence processing efficient and generic. To illustrate our approach, we describe two multi-perspective studio applications, and discuss performance and scalability results.
Our current research effort aims at building a filter based post-OCR accuracy boost system that will combine different post-OCR correction filters to improve the OCR accuracy better than each individual filter can. In this paper we focus on a Hidden Markov Model (HMM) based accuracy booster modeling OCR engine noise generation as a two-layer stochastic process. We employ a commercial spellchecker both as another error correction filter and as a base line for accuracy boost comparison. We demonstrate the versatility of our approach in experiments with documents in English and Arabic.
We introduce a generic and efficient method for 2D and 3D shape estimation via density Þelds. Our method models shape as a density map and uses the notion of density to Þt a model to a rapidly computed occupancy map of the foreground object. We show how to utilize hierarchical (pyramid-like) object segmentation data to regularize a hierarchical model Þtting. With primary focus on estimating 3D shapes of non-rigid articulated objects such as human bodies, we illustrate our approach with examples of efficient model Þtting to 3D occupancy maps of human Þgures. We also discuss a number of extensions of our method to applications involving non-rigid object tracking and movement analysis.
This report explores the latest advances in the field of digital document recognition. With the focus on printed document imagery, we discuss the major developments in optical character recognition (OCR) and document image enhancement/restoration in application to Latin and non-Latin scripts. In addition, we review and discuss the available technologies for hand-written document recognition. In this report, we also provide some company-accumulated benchmark results on available OCR engines.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.