Image compression method

Image analysis – Image compression or coding – Quantization

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C382S238000

Reexamination Certificate

active

06438267

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image coding apparatus and image coding method, a recording medium, an image decoding apparatus, an image decoding method, a learning apparatus, and a learning method. More particularly, the present invention relates to an image coding apparatus for coding an image by, for example, subsampling it so that a decoded image which is almost the same as the original image can be obtained, and an image coding method, a recording medium, an image decoding apparatus, an image decoding method, a learning apparatus, and a learning method.
2. Description of the Related Art
Various methods have been put forward as methods for coding an image. For example, one such method is a method for compressing and coding an image by subsampling the pixels thereof.
However, when an image (image of a high-order hierarchy) compressed by subsampling it in this manner is expanded by simple interpolation (for example, the interpolated value of peripheral pixels), the resolution of the decoded image obtained is deteriorated.
The causes for the deterioration in resolution of the decoded image in this manner are as follows. Firstly, a subsampled image does not contain high-frequency components which are in the original image. Secondly, the pixel values of the pixels which form the image after being subsampled are not necessarily appropriate for reconstructing the original image.
The present invention has been achieved in view of such circumstances. An object of the present invention is to make it possible to subsample an image into compressed code quickly so that a decoded image which is the same (almost the same) as the original image can be obtained.
SUMMARY OF THE INVENTION
As the present invention sets out to solve these problems, the present invention is capable of producing an optimum compressed data in such a manner that decoded images almost the same in quality as the original image can be obtained.
According to one aspect of the present invention, an image encoding method compresses an original image data to a fewer number of pixels by performing a class classification to the compressed data and determining a class result corresponding to the compressed data, generating prediction data of the original data in accordance with the class result, generating prediction error information, estimating error curve information representing a relationship between the compressed data and the prediction error information, and determining an optimum compressed data in accordance with the error curve information.
According to another aspect of the present invention, an image encoding apparatus includes means for reducing a number of pixels of an original image and generating a compressed data and means for performing a class classification to the compressed data and determining a class result corresponding to the compressed data. Prediction means generates prediction data of the original data in accordance with the class result. Prediction error information means generates prediction error information representing a difference between the prediction data and the original data. Estimation means estimates error curve information representing a relationship between the compressed data and the prediction error information. Determination means determines an optimum compressed data in accordance with the error curve information.
According to yet another aspect of the present invention, the prediction data generating means includes a memory in which prediction coefficients for each clang are stored and output corresponding to the class result therefrom, the prediction coefficients for each class produced by learning in advance, using a number of training image, and prediction means generating prediction data of the original data in accordance with the read prediction coefficients and the compressed data.
According to a further aspect of the present invention, a recording medium recorded with an encoding program is prepared by reducing a number of pixels of an original image and generating a compressed data, performing a class classification to the compressed data and determining a class result corresponding to the compressed data, generating prediction data of the original data in accordance with the class result, generating prediction error information representing a difference between the prediction data and the original data, estimating error curve information representing a relationship between the compressed data and the prediction error information, and determining an optimum compressed data in accordance with the error curve information.
These and other aspects of the invention will become apparent when considered with reference to the following drawings.


REFERENCES:
patent: 4703352 (1987-10-01), Kondo
patent: 5122873 (1992-06-01), Golin
patent: 5193003 (1993-03-01), Kondo
patent: 5200962 (1993-04-01), Kao et al.
patent: 5293230 (1994-03-01), Golin
patent: 5331414 (1994-07-01), Golin
patent: 5349385 (1994-09-01), Glenn
patent: 5384869 (1995-01-01), Wilkinson et al.
patent: 5469216 (1995-11-01), Takahashi et al.
patent: 5488618 (1996-01-01), Kondo et al.
patent: 5495297 (1996-02-01), Fujimori et al.
patent: 5504535 (1996-04-01), Abe
patent: 5517245 (1996-05-01), Kondo et al.
patent: 5528606 (1996-06-01), Kondo et al.
patent: 5553160 (1996-09-01), Dawson
patent: 5598214 (1997-01-01), Kondo et al.
patent: 5610658 (1997-03-01), Uchida et al.
patent: 5625712 (1997-04-01), Schoenzeit et al.
patent: 5663764 (1997-09-01), Kondo et al.
patent: 5666164 (1997-09-01), Kondo et al.
patent: 5680225 (1997-10-01), Hirabayashi et al.
patent: 5687257 (1997-11-01), Paik et al.
patent: 5734433 (1998-03-01), Kondo et al.
patent: 5739873 (1998-04-01), Kondo
patent: 5754702 (1998-05-01), Simpson
patent: 5796442 (1998-08-01), Gove et al.
patent: 5812699 (1998-09-01), Zhu et al.
patent: 5825313 (1998-10-01), Kondo et al.
patent: 5867593 (1999-02-01), Fukuda et al.
patent: 5870434 (1999-02-01), Kondo et al.
patent: 5880784 (1999-03-01), Lillevold
patent: 5912708 (1999-06-01), Kondo et al.
patent: 5930394 (1999-07-01), Kondo et al.
patent: 5960116 (1999-09-01), Kajiwara
patent: 5966179 (1999-10-01), Kondo et al.
patent: 5969764 (1999-10-01), Sun et al.
patent: 6016164 (2000-01-01), Kawaguchi et al.
patent: 6233358 (2001-05-01), Acharya
patent: 0293041 (1988-11-01), None
patent: 0635978 (1995-01-01), None
patent: 3-53778 (1991-03-01), None
patent: WO 91/20159 (1991-12-01), None
patent: WO 92/07445 (1992-04-01), None
patent: WO 93/13624 (1993-07-01), None
Patent Abstracts of Japan, JP 7-147681, Published Jun. 6, 1995, Sony Corporation.
I. Chang Et Al., “Adaptive Subsampling JPEG Image Coding,” IEEE Jun. 1995, pp 264-265.
Application No. 08/893,134, Jul. 15, 1997.
Application No. 08/892,570, Jul. 15, 1997.
Application No. 08/893,022, Jul. 15, 1997.
Application No. 08/893,216, Jul. 15, 1997
Application No. 08/893,202, Jul. 15, 1997.

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Image compression 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 compression method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image compression method will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-2973966

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.