Image analysis – Pattern recognition – Classification
Reexamination Certificate
2000-04-28
2001-07-17
Boudreau, Leo (Department: 2721)
Image analysis
Pattern recognition
Classification
C382S232000, C382S235000, C382S238000, C382S239000, C382S243000, C382S249000, C348S390100, C348S394100, C348S420100
Reexamination Certificate
active
06263105
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus and method and, in particular, to an image processing apparatus and method which classify an image as one of several classes in accordance with its property and subject it to a process corresponding to its class.
For example, among various methods for performing image processing such as converting a standard or low resolution image (hereinafter referred to as an SD image where appropriate) into a high resolution image (hereinafter referred to as an HD image where appropriate) is a method which classifies each block that constitutes an SD image as one of predetermined classes in accordance with its property and subjects it to image processing corresponding to its class.
For example, it is now assumed that, as shown in
Fig. 1A
, a block of 2×2 pixels is formed by a certain subject pixel and three pixels adjacent to it and each pixel is expressed by one bit (that is, it has a level of 0 or 1). In this case, as shown in
FIG. 1B
, blocks of 4 (2×2) pixels can be classified into 16 (=(2
1
)
4
) patterns based on the distribution of the levels of the respective pixels. The classification means this type of division into patterns, and the classification enables each block to be subjected to image processing suitable for its class, i.e., its image property.
However, image properties (features) are not represented by only the level distribution of pixels. It would therefore be considered that an image may be processed more properly by classifying it by using some other feature quantity.
SUMMARY OF THE INVENTION
The present invention has been made in view of the above circumstances, and an object of the invention is therefore to make it possible to process an image more properly in accordance with its properties.
An image processing apparatus according to the invention is characterized by calculating means for calculating self-similarity of an image based on a compressed image, and classifying means for determining a class of the image from among a plurality of classes based on the calculated self-similarity.
An image processing method according to the invention is characterized by the steps of calculating self-similarity of an image based on a compressed image, and determining a class of the image from among a plurality of classes based on the calculated self-similarity.
More specifically, the invention is summarized as follows by using, in parentheses, exemplary terms appearing in the embodiments.
An image processing apparatus for classifying an image into a plurality of classes in accordance with a property thereof and for executing processes corresponding to the respective classes, comprising generating means (a decimation circuit
2
and a reduction circuit
3
in
FIG. 2
) for generating, from the image, a compressed image having a smaller number of pixels than the image; calculating means (similarily calculation sections
21
1
-
21
4
in
FIG. 5
) for calculating self-similarity of the image based on the compressed image; classifying means (a maximum similarity direction judgment section
22
in
FIG. 5 and a
pattern classification section
33
in
FIG. 6
) for determining a class of the image from among the plurality of classes based on the calculated self-similarity; and processing means (a prediction circuit
6
in
FIG. 2
) for executing a process corresponding to the determined class on the image.
The above image processing apparatus may be constructed such that the generating means generates, as the compressed image, a decimated image obtained by decimating pixels of the image and a reduced image obtained by eliminating peripheral pixels of the image, and that the calculating means comprises first extracting means (a pixels extraction section
26
C in
FIG. 5
) for extracting pixels that constitute the decimated image and arranged in a given direction; second extracting means (a pixels extraction section
26
B in
FIG. 5
) for extracting pixels that constitute the reduced image and arranged in the given direction; and operating means (a norm calculation section
27
in
FIG. 5
) for calculating, as the self-similarity in the given direction, a norm of vectors having, as components, pixel values of the pixels extracted by the first and second extracting means, respectively.
The above image processing apparatus may further comprise detecting means (a maximum similarity direction pixels extraction section
31
in
FIG. 6
) for detecting a pattern of pixel values of pixels arranged in a direction in which the self-similarity calculated by the calculating means is highest, and the classifying means may classify the image as a class corresponding to the detected pattern.
Naturally, the above specific circuits and sections cited in parentheses are mere examples and should not be construed as limiting the corresponding means.
REFERENCES:
patent: 5311308 (1994-05-01), Bernard
patent: 5485554 (1996-01-01), Lowitz et al.
patent: 5555322 (1996-09-01), Terai et al.
patent: 5654761 (1997-08-01), Jung
patent: 5671294 (1997-09-01), Rogers et al.
patent: 5701369 (1997-12-01), Moon et al.
patent: 5754676 (1998-05-01), Komiya et al.
patent: 5764305 (1998-06-01), Kondo
patent: 5781238 (1998-07-01), Kondo et al.
patent: 5787201 (1998-07-01), Nelson et al.
patent: 5862263 (1999-01-01), Kim et al.
patent: 5892847 (1999-04-01), Johnson
Kiselyov, et al. Self-Similarity of the Multiresolutional Image/Video Decomposition: Smart Expansion as Compression of Still and Moving Pictures,IEEE,4/94.
Liu, et al. “Fractal Description and Classification of Breast Tumors”, IEEE, 4/91.
Chen, et al. “Fractal Feature Analysis and Classification in Medical Imaging”, IEEE, 2/89.
Kondo Tetsujiro
Nakaya Hideo
Boudreau Leo
Chang Dexter T.
Frommer William S.
Frommer & Lawrence & Haug LLP
Mariam Daniel G.
LandOfFree
Image processing apparatus and method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing apparatus and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing apparatus and method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2440551