Red-Eyes Removal through Cluster-Based Boosting on Gray Codes
<p/> <p>Since the large diffusion of digital camera and mobile devices with embedded camera and flashgun, the redeyes artifacts have de facto become a critical problem. The technique herein described makes use of three main steps to identify and remove red eyes. First, red-eye candidates...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
SpringerOpen
2010-01-01
|
Series: | EURASIP Journal on Image and Video Processing |
Online Access: | http://jivp.eurasipjournals.com/content/2010/909043 |
id |
doaj-823f283d9c39464eb1ef59bf589a2c84 |
---|---|
record_format |
Article |
spelling |
doaj-823f283d9c39464eb1ef59bf589a2c842020-11-24T22:10:05ZengSpringerOpenEURASIP Journal on Image and Video Processing1687-51761687-52812010-01-0120101909043Red-Eyes Removal through Cluster-Based Boosting on Gray CodesGuarnera MirkoMessina GiuseppeBattiato SebastianoFarinella GiovanniMariaRavì Daniele<p/> <p>Since the large diffusion of digital camera and mobile devices with embedded camera and flashgun, the redeyes artifacts have de facto become a critical problem. The technique herein described makes use of three main steps to identify and remove red eyes. First, red-eye candidates are extracted from the input image by using an image filtering pipeline. A set of classifiers is then learned on gray code features extracted in the clustered patches space and hence employed to distinguish between eyes and non-eyes patches. Specifically, for each cluster the gray code of the red-eyes candidate is computed and some discriminative gray code bits are selected employing a boosting approach. The selected gray code bits are used during the classification to discriminate between <it>eye</it> versus <it>non-eye</it> patches. Once red-eyes are detected, artifacts are removed through desaturation and brightness reduction. Experimental results on a large dataset of images demonstrate the effectiveness of the proposed pipeline that outperforms other existing solutions in terms of hit rates maximization, false positives reduction, and quality measure.</p>http://jivp.eurasipjournals.com/content/2010/909043 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Guarnera Mirko Messina Giuseppe Battiato Sebastiano Farinella GiovanniMaria Ravì Daniele |
spellingShingle |
Guarnera Mirko Messina Giuseppe Battiato Sebastiano Farinella GiovanniMaria Ravì Daniele Red-Eyes Removal through Cluster-Based Boosting on Gray Codes EURASIP Journal on Image and Video Processing |
author_facet |
Guarnera Mirko Messina Giuseppe Battiato Sebastiano Farinella GiovanniMaria Ravì Daniele |
author_sort |
Guarnera Mirko |
title |
Red-Eyes Removal through Cluster-Based Boosting on Gray Codes |
title_short |
Red-Eyes Removal through Cluster-Based Boosting on Gray Codes |
title_full |
Red-Eyes Removal through Cluster-Based Boosting on Gray Codes |
title_fullStr |
Red-Eyes Removal through Cluster-Based Boosting on Gray Codes |
title_full_unstemmed |
Red-Eyes Removal through Cluster-Based Boosting on Gray Codes |
title_sort |
red-eyes removal through cluster-based boosting on gray codes |
publisher |
SpringerOpen |
series |
EURASIP Journal on Image and Video Processing |
issn |
1687-5176 1687-5281 |
publishDate |
2010-01-01 |
description |
<p/> <p>Since the large diffusion of digital camera and mobile devices with embedded camera and flashgun, the redeyes artifacts have de facto become a critical problem. The technique herein described makes use of three main steps to identify and remove red eyes. First, red-eye candidates are extracted from the input image by using an image filtering pipeline. A set of classifiers is then learned on gray code features extracted in the clustered patches space and hence employed to distinguish between eyes and non-eyes patches. Specifically, for each cluster the gray code of the red-eyes candidate is computed and some discriminative gray code bits are selected employing a boosting approach. The selected gray code bits are used during the classification to discriminate between <it>eye</it> versus <it>non-eye</it> patches. Once red-eyes are detected, artifacts are removed through desaturation and brightness reduction. Experimental results on a large dataset of images demonstrate the effectiveness of the proposed pipeline that outperforms other existing solutions in terms of hit rates maximization, false positives reduction, and quality measure.</p> |
url |
http://jivp.eurasipjournals.com/content/2010/909043 |
work_keys_str_mv |
AT guarneramirko redeyesremovalthroughclusterbasedboostingongraycodes AT messinagiuseppe redeyesremovalthroughclusterbasedboostingongraycodes AT battiatosebastiano redeyesremovalthroughclusterbasedboostingongraycodes AT farinellagiovannimaria redeyesremovalthroughclusterbasedboostingongraycodes AT rav236daniele redeyesremovalthroughclusterbasedboostingongraycodes |
_version_ |
1725809459361480704 |