Adaptive data transmission on browsing of scanned documents using JPEG2000

Abstract

In this paper we propose an adaptive stream data transmission mechanism on transmitting and browsing the high-resolution scanned document images. We focus on a content delivery system for easily transmitting and browsing a large-scale of document image database such as newspapers, magazines, and in-house documents. We discuss the problems that remain for stream data transmission and describe the schemes of characteristic functions being used in our system. The JPEG2000 format is used to compress the scanned documents' images and moreover, layout analysis is employed to get the structure of each page. Our mechanism enables a web client to browse the such document images with the reality of flipping through a real paper document. Based on the characteristics of JPEG2000 format and the information of each page layout analysis result the web client can select an interested region to show detail even if on a small screen. We present a prototype of the implementation system and realize the high-speed document browsing in mobile web environment.