Image analysis – Image compression or coding – Adaptive coding
Reexamination Certificate
2000-10-27
2004-09-14
Couso, Jose L. (Department: 2621)
Image analysis
Image compression or coding
Adaptive coding
Reexamination Certificate
active
06792152
ABSTRACT:
TECHNICAL FIELD
The present invention relates to an image coding method, an image coding and decoding method, an image coding apparatus, and an image recording and reproduction apparatus and, more particularly, to those used for a system which codes a video signal with high efficiency and records it on an image recording medium such as an optical disk.
BACKGROUND ART
As a conventional system for coding a video signal with high efficiency and recording it, there has been known an image coding apparatus disclosed in Japanese Published Patent Application No. Hei. 9-18872.
FIG. 20
is a simplified block diagram of the conventional image coding apparatus.
FIGS. 21
to
23
are diagrams of characteristics for deciding a quantization step by the conventional image coding apparatus. In each of
FIGS. 21
to
23
, the abscissa shows the size of each image feature data, and the ordinate shows the coefficient of quantization characteristic. The larger the coefficient is, the smaller the quantization step is. That is, the number of bits to be given is increased, and the quantity of generated codes is increased.
The construction and operation will be described with reference to FIG.
20
. As shown in
FIG. 20
, the conventional image coding apparatus comprises an average luminance level extractor
2001
for extracting the average of the luminance signal for every predetermined period; an average chrominance level extractor
2002
for extracting the averages of the color-difference signals for every predetermined period; a luminance variance extractor
2003
for extracting the variance of the luminance signal for every predetermined period, a quantization step decider
2004
, and an encoder
2005
.
An image signal is input to the average luminance level extractor
2001
, the average chrominance level extractor
2002
, the luminance variance extractor
2003
, and the encoder
2005
. In the average luminance level extractor
2001
, the average chrominance level extractor
2002
, and the luminance variance extractor
2003
, the respective image feature data are extracted from the input image signal over a predetermined period, and the extracted data are input to the quantization step decider
2004
. The quantization step decider
2004
decides a quantization step according to at least one of these image feature data, and outputs the result to the encoder
2005
. The encoder
2005
encodes the input image signal using the quantization step decided for every predetermined period by the quantization step decider
2004
, and outputs it.
The quantization step decider
2004
operates as follows. When the average luminance level extracted by the average luminance level extractor
2001
is relatively small, the decider
2004
decreases the quantization step according to the characteristics shown in FIG.
21
. Between two color-difference signals Cb and Cr extracted by the average chrominance level extractor
2002
, when the average level of the red-base Cr signal is larger than that of the Cb signal, the decider
2004
decreases the quantization step according to the characteristics shown in FIG.
22
. When the luminance variance extracted by the luminance variance extractor
2003
is relatively small, the decider
2004
decreases the quantization step according to the characteristics shown in FIG.
23
. The coefficients at the ordinates of
FIGS. 21
to
23
are weights on the quantization step. The larger the coefficient is, the smaller the quantization step is.
In this way, the quantization step is decided adaptively to the human visual characteristics, followed by coding.
As described above, in the conventional image coding apparatus, the quantization step is adaptively decided to the human visual characteristics by using the rough feature data of the image signal for every predetermined period, and then the image signal is coded. Therefore, when coding is performed according to the extracted image feature data to reduce image degradation, the quantity of generated codes is increased as a whole. On the contrary, when the quantity of generated codes is reduced, image degradation occurs over the entire image displayed in the predetermined period.
The present invention is made to solve the above-described problems and has for its object to provide an image coding method, an image coding and decoding apparatus, an image coding apparatus, and an image recording and reproduction apparatus, which are adapted to the human visual characteristics on the entire image, and reduce degradation of image quality while minimizing the influence on the quantity of generated codes.
DISCLOSURE OF INVENTION
According to one aspect of the present invention, there is provided an image coding method comprising: dividing an input image signal into local regions; deciding a quantization characteristic for each local region of the input image signal on the basis of the frequency distribution of image feature data of each local region over a predetermined period; and coding the image signal. Therefore, the quantity of generated codes in each local region can be precisely controlled, and occurrence of coding noise can be controlled, by utilizing the statistical data of the whole image.
According to a second aspect of the present invention, there is provided an image coding method comprising: dividing an input image signal into local regions; deciding a filter characteristic for each local region of the input image signal on the basis of image feature data of each local region; performing adaptive filtering; and coding the image signal. Since the noise component is locally suppressed or the frequency band is controlled by the adaptive filtering before coding, occurrence of coding noise and increase in the quantity of generated codes can be suppressed.
According to a third aspect of the present invention, there is provided an image coding method comprising: dividing an input image signal into local regions; deciding a filter characteristic for each local region of the input image signal on the basis of the frequency distribution of image feature data of each local region over a predetermined period; performing adaptive filtering; and coding the image signal. Since the noise component is locally suppressed or the frequency band is controlled before coding by the adaptive filtering utilizing the statistical data of the whole image, occurrence of coding noise and increase in the quantity of generated codes can be suppressed.
According to a fourth aspect of the present invention, in the image coding method of the first aspect, employed as an input image signal is a signal obtained by dividing an input image signal into local regions, deciding a filter characteristic for each local region of the input image signal on the basis of image feature data of each local region, and subjecting the image signal to adaptive filtering. Therefore, occurrence of coding noise and increase in the quantity of generated codes can be suppressed for each local region. Further, the quantity of generated codes can be precisely controlled for each local region, and occurrence of coding noise can be controlled.
According to a fifth aspect of the present invention, in the image coding method of the first aspect, employed as an input image signal is a signal obtained by dividing an input image signal into local regions, deciding a filter characteristic for each local region of the input image signal on the basis of the frequency distribution of image feature data of each local region over a predetermined period, and subjecting the image signal to adaptive filtering. Therefore, occurrence of coding noise and increase in the quantity of generated codes can be suppressed for each local region by utilizing the statistical data of the whole image. Further, the quantity of generated codes can be precisely controlled for each local region, and occurrence of coding noise can be controlled.
According to a sixth aspect of the present invention, there is provided an image coding method comprising: dividing an input image signal into plural regions; deciding a filter c
Fukuda Hideki
Nakamura Kazuhiko
Shibata Hideaki
Couso Jose L.
Matsushita Electric - Industrial Co., Ltd.
Wenderoth , Lind & Ponack, L.L.P.
LandOfFree
Image coding method, image coding/decoding method, image... 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 coding method, image coding/decoding method, image..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image coding method, image coding/decoding method, image... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3191747