Image-based applications can save time and space by operating on compressed data. The problem is that most mid- and high-level image operations, such as object recognition, are formulated as sequences of operations in the image domain. Such methods need direct access to pixel information as a starting point, but the pixel information in a compressed image stream is not immediately accessible. In this paper we show how to perform object recognition directly on compressed images (JPEG) and index frames from video streams (MPEG I-frames) without recovering explicit pixel information. The approach uses eigenvectors constructed from compressed image data. Our performance results show that a five-fold speedup can be gained by using compressed data.
|Number of pages||16|
|Journal||Image and Vision Computing|
|State||Published - Apr 27 1998|
Bibliographical noteFunding Information:
The authors would like to thank the MIT Media Laboratory, the Olivetti Research Laboratory, and the Machine Vision group at the University of Essex for making their face recognition data and code publicly available. We thank Zhaojun Bai, Tom Hayden, Cid Srinivasan, Raphael Finkel and Mirek Truszczynski (University of Kentucky) for comments and valuable discussions. Finally, we gratefully acknowledge the support of the National Science Foundation for this project under grant number IRI-9308415.
- Compressed images
- Object recognition
ASJC Scopus subject areas
- Signal Processing
- Computer Vision and Pattern Recognition