| Evaluating color descriptors for object and scene recognition. | |
| | |
MedLine Citation:
|
PMID: 20634554 Owner: NLM Status: MEDLINE |
Abstract/OtherAbstract:
|
Image category recognition is important to access visual information on the level of objects and scene types. So far, intensity-based descriptors have been widely used for feature extraction at salient points. To increase illumination invariance and discriminative power, color descriptors have been proposed. Because many different descriptors exist, a structured overview is required of color invariant descriptors in the context of image category recognition. Therefore, this paper studies the invariance properties and the distinctiveness of color descriptors (software to compute the color descriptors from this paper is available from http://www.colordescriptors.com) in a structured way. The analytical invariance properties of color descriptors are explored, using a taxonomy based on invariance properties with respect to photometric transformations, and tested experimentally using a data set with known illumination conditions. In addition, the distinctiveness of color descriptors is assessed experimentally using two benchmarks, one from the image domain and one from the video domain. From the theoretical and experimental results, it can be derived that invariance to light intensity changes and light color changes affects category recognition. The results further reveal that, for light intensity shifts, the usefulness of invariance is category-specific. Overall, when choosing a single descriptor and no prior knowledge about the data set and object and scene categories is available, the OpponentSIFT is recommended. Furthermore, a combined set of color descriptors outperforms intensity-based SIFT and improves category recognition by 8 percent on the PASCAL VOC 2007 and by 7 percent on the Mediamill Challenge. |
| | |
Authors:
|
Koen E A van de Sande; Theo Gevers; Cees G M Snoek |
Related Documents
:
|
3878274 - Color mixture thresholds measured on a color television--a new method for analysis, cla... 7644324 - Abrupt luminance change pops out; abrupt color change does not. 18352924 - Fabricating a better mouthguard. part ii: the effect of color on adaptation and fit. 12207974 - Depth from subjective color and apparent motion. 20448334 - Behavior of toads, bufo bufo, in a dynamic environment. 22563314 - A wavelet-based neural model to optimize and read out a temporal population code. |
Publication Detail:
|
Type: Journal Article; Research Support, Non-U.S. Gov't |
Journal Detail:
|
Title: IEEE transactions on pattern analysis and machine intelligence Volume: 32 ISSN: 1939-3539 ISO Abbreviation: IEEE Trans Pattern Anal Mach Intell Publication Date: 2010 Sep |
Date Detail:
|
Created Date: 2010-07-16 Completed Date: 2010-12-27 Revised Date: - |
Medline Journal Info:
|
Nlm Unique ID: 9885960 Medline TA: IEEE Trans Pattern Anal Mach Intell Country: United States |
Other Details:
|
Languages: eng Pagination: 1582-96 Citation Subset: IM |
Affiliation:
|
Informatics Institute, University of Amsterdam, Science Park 107, 1098 XG Amsterdam, The Netherlands. ksande@uva.nl |
Export Citation:
|
APA/MLA Format Download EndNote Download BibTex |
| MeSH Terms | |
Descriptor/Qualifier:
|
Algorithms* Artificial Intelligence* Color* Colorimetry / methods* Image Enhancement / methods Image Interpretation, Computer-Assisted / methods* Imaging, Three-Dimensional / methods* Pattern Recognition, Automated / methods* Reproducibility of Results Sensitivity and Specificity |
From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine
Previous Document: Efficient sequential correspondence selection by cosegmentation.
Next Document: Geometric feature extraction by a multimarked point process.