Saliency detection in the compressed domain for adaptive image retargeting
Saliency detection plays important roles in many image processing applications, such as regions of interest extraction and image resizing. Existing saliency detection models are built in the uncompressed domain. Since most images over Internet are typically stored in the compressed domain such as jo...
Saved in:
Main Authors: | , , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
2013
|
Online Access: | https://hdl.handle.net/10356/98840 http://hdl.handle.net/10220/13484 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-98840 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-988402020-05-28T07:18:56Z Saliency detection in the compressed domain for adaptive image retargeting Fang, Yuming Chen, Zhenzhong Lin, Weisi Lin, Chia-Wen School of Computer Engineering School of Electrical and Electronic Engineering Saliency detection plays important roles in many image processing applications, such as regions of interest extraction and image resizing. Existing saliency detection models are built in the uncompressed domain. Since most images over Internet are typically stored in the compressed domain such as joint photographic experts group (JPEG), we propose a novel saliency detection model in the compressed domain in this paper. The intensity, color, and texture features of the image are extracted from discrete cosine transform (DCT) coefficients in the JPEG bit-stream. Saliency value of each DCT block is obtained based on the Hausdorff distance calculation and feature map fusion. Based on the proposed saliency detection model, we further design an adaptive image retargeting algorithm in the compressed domain. The proposed image retargeting algorithm utilizes multioperator operation comprised of the block-based seam carving and the image scaling to resize images. A new definition of texture homogeneity is given to determine the amount of removal block-based seams. Thanks to the directly derived accurate saliency information from the compressed domain, the proposed image retargeting algorithm effectively preserves the visually important regions for images, efficiently removes the less crucial regions, and therefore significantly outperforms the relevant state-of-the-art algorithms, as demonstrated with the in-depth analysis in the extensive experiments. 2013-09-16T07:08:21Z 2019-12-06T20:00:11Z 2013-09-16T07:08:21Z 2019-12-06T20:00:11Z 2012 2012 Journal Article 1057-7149 https://hdl.handle.net/10356/98840 http://hdl.handle.net/10220/13484 10.1109/TIP.2012.2199126 en IEEE transactions on image processing © 2012 IEEE |
institution |
Nanyang Technological University |
building |
NTU Library |
country |
Singapore |
collection |
DR-NTU |
language |
English |
description |
Saliency detection plays important roles in many image processing applications, such as regions of interest extraction and image resizing. Existing saliency detection models are built in the uncompressed domain. Since most images over Internet are typically stored in the compressed domain such as joint photographic experts group (JPEG), we propose a novel saliency detection model in the compressed domain in this paper. The intensity, color, and texture features of the image are extracted from discrete cosine transform (DCT) coefficients in the JPEG bit-stream. Saliency value of each DCT block is obtained based on the Hausdorff distance calculation and feature map fusion. Based on the proposed saliency detection model, we further design an adaptive image retargeting algorithm in the compressed domain. The proposed image retargeting algorithm utilizes multioperator operation comprised of the block-based seam carving and the image scaling to resize images. A new definition of texture homogeneity is given to determine the amount of removal block-based seams. Thanks to the directly derived accurate saliency information from the compressed domain, the proposed image retargeting algorithm effectively preserves the visually important regions for images, efficiently removes the less crucial regions, and therefore significantly outperforms the relevant state-of-the-art algorithms, as demonstrated with the in-depth analysis in the extensive experiments. |
author2 |
School of Computer Engineering |
author_facet |
School of Computer Engineering Fang, Yuming Chen, Zhenzhong Lin, Weisi Lin, Chia-Wen |
format |
Article |
author |
Fang, Yuming Chen, Zhenzhong Lin, Weisi Lin, Chia-Wen |
spellingShingle |
Fang, Yuming Chen, Zhenzhong Lin, Weisi Lin, Chia-Wen Saliency detection in the compressed domain for adaptive image retargeting |
author_sort |
Fang, Yuming |
title |
Saliency detection in the compressed domain for adaptive image retargeting |
title_short |
Saliency detection in the compressed domain for adaptive image retargeting |
title_full |
Saliency detection in the compressed domain for adaptive image retargeting |
title_fullStr |
Saliency detection in the compressed domain for adaptive image retargeting |
title_full_unstemmed |
Saliency detection in the compressed domain for adaptive image retargeting |
title_sort |
saliency detection in the compressed domain for adaptive image retargeting |
publishDate |
2013 |
url |
https://hdl.handle.net/10356/98840 http://hdl.handle.net/10220/13484 |
_version_ |
1681059701433827328 |