Image analysis – Color image processing – Image segmentation using color
Reexamination Certificate
2001-02-20
2004-08-24
Wu, Jingge (Department: 2623)
Image analysis
Color image processing
Image segmentation using color
Reexamination Certificate
active
06782126
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to document image analysis and, more particularly, to segmentation of an image using a method which combines distance in the feature space with distance in the spatial domain to produce regions of near-uniform color.
2. Background Description
In many applications, such as document image analysis, and analysis of digital images, an important processing stage is that of segmentation of the image into regions of near-uniform color. The results of this stage are used for further analysis, such as a determination of the number of colors present in the image, identification of regions with specific color and an analysis of geometric features of regions with uniform color. Applying color cluster analysis to the entire image is a time-consuming process, and also results in too many clusters being identified.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide an improved method for the complete segmentation of an image into regions of near-uniform color.
According to the invention, there is provided a method which segments the image into blocks of size M×N (say, 128×128) and applies color clustering within these blocks. This is done with a sequential color clustering algorithm. Subsequent to this, the blocks that have similar color characteristics are identified. Finally, a connected component analysis in the spatial domain is performed to identify blocks that are both similar in color characteristics as well as spatially contiguous. Thus, our invention combines distance in the feature space with distance in the spatial domain. This results in spatially compact color regions which have similar color information.
The result of applying this technique is a complete segmentation of the image according to regions of near-uniform color. This can then be used for performing operations like optical character recognition, or adaptive compression of the image on the number of colors present.
REFERENCES:
patent: 6128407 (2000-10-01), Inoue et al.
patent: 6381363 (2002-04-01), Murching
patent: 6483940 (2002-11-01), Wang
patent: 6529202 (2003-03-01), Wu
Kuo “Unsupervised color image segmentation for content based application” IEEE International Confference on Multimedia and Expo, 2000 vol. 3, pp. 1515-1518, Aug. 2000.*
Article titled “Unsupervised Color Image Segmentation for Content Based Application” Chumg Hui Kuo and Ahmed H. Tewfik; dated 2000.
Article titled “Sigma Filter Based Unsupervised Color Image Segmentation” Chumg Hui Kuo and Ahmed H. Tewfik; dated 2000.
Pattern Recognition, by S. Theodoridis and K. Koutroumbas, Academic Press, 1999, pp. 387-392.
Pattern Recognition by S. Theodoridis and K. Koutroumbas, Academic Press, 1999, pp. 497-502.
Algorithms for Clustering Data, by A.K. Jain and R. C. Dubes, Prentice Hall, 1988, pp. 120-124.
Robot Vision, by B. K. P. Horn, MIT Press, 1986, pp. 65-71.
Chung-Hui Kuo
Rao Ravishankar
Thompson Gerhard R.
Kaufman Stephen C.
Whitham Curtis & Christofferson, P.C.
Wu Jingge
LandOfFree
Method for combining feature distance with spatial distance... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for combining feature distance with spatial distance..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for combining feature distance with spatial distance... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3272975