Image analysis – Image compression or coding – Adaptive coding
Reexamination Certificate
2000-06-08
2003-07-29
Couso, Jose L. (Department: 2621)
Image analysis
Image compression or coding
Adaptive coding
Reexamination Certificate
active
06600837
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image compression controller applied, for example, for a digital still camera system and compressing image data to generate an image file of a size of not more than a set size and an image pickup system using the same.
2. Description of the Related Art
Compression based on for example the Joint Photographic Experts Group (JPEG) standard of still images may be roughly divided into generally the four processes in order: discrete cosine transform (DCT) processing; quantization; entropy coding; and JPEG file formatting.
Here, the DCT circuit performs orthogonal transform processing applying two-dimensional DCT on original image data to remove the spatial relationship, focusing the energy on some coefficients, and compressing and coding so-called redundancy.
The quantization approximates the data with the compressed redundancy at a discrete level.
The entropy coding assigns codes for every quantization output so that the average code length becomes short.
The JPEG file formatting produces a JPEG file of a size of not more than a set size from the entropy coded data.
The final JPEG data is obtained by all these processings, so the data size is found after completing the processing.
Also, the JPEG data size, or the compression rate, changes in accordance with the control of a quantization coefficient in the quantization stage.
Controlling the JPEG data size to be a desired size becomes possible by repeatedly performing the above processing while changing the quantization coefficient.
However, the method of controlling the JPEG data size to be a desired size by changing the quantization coefficient requires a long processing time.
Especially, in a digital still camera system etc., since the capacity of the storage medium for storing the JPEG file is limited, when compressing a photographed still image, control is required so that the final JPEG data does not become larger than a certain size.
Accordingly, a digital still camera system etc. is provided with an image compression controller for changing the quantization coefficient by for example an image quality mode so as to control the size of the JPEG data.
FIG. 13
is a block diagram of an example of the configuration of this type of image compression controller.
An image compression controller
10
comprises a DCT circuit
11
, a quantization circuit
12
, an entropy coding circuit
13
, a size measurement circuit
14
, a quantization coefficient control circuit
15
, a quantization table
16
stored in a storage circuit, and a JPEG file maker
17
.
In the DCT circuit
11
of the image compression controller
10
, as shown in
FIG. 14
, a discrete cosine transform on is performed with respect to the input original image data and the result is output as data D
11
to the quantization circuit
12
(ST
1
, ST
2
).
In the quantization circuit
12
, the data D
11
based on a quantization coefficient S
15
given by the quantization coefficient control circuit
15
is quantized (ST
3
) and the result is output as data D
12
to the entropy coding circuit
13
.
At this time, the quantization coefficient given by the quantization coefficient control circuit
15
is given by selecting a coefficient having a low compression rate at the time of initialization from the quantization table
16
.
In the entropy coding circuit
13
, codes are assigned to the quantization data D
12
so that the average code length becomes short (ST
4
) and the result is output as data D
13
to the size measurement circuit
14
and the JPEG file maker
17
.
In the size measurement circuit
14
, the size of the entropy coded data D
13
(ST
5
) is measured and the measured size is output as a signal S
14
to the quantization coefficient control circuit
15
.
In the quantization coefficient control circuit
15
, the signal S
14
is received and the judgement is performed whether the entropy coded data size is less than a set predetermined size (ST
6
).
When the quantization coefficient control circuit
15
judges that the entropy coded size is larger than the set size at step ST
6
, a signal S
15
a
indicating the judgement is output to the JPEG file maker
17
, selected a quantization coefficient which gives a little higher compression rate than that at first from the quantization table
16
, and this is given to the quantization circuit
12
(ST
7
). At this time the JPEG file maker
17
does not make a file.
The processing of steps ST
1
to ST
7
is then repeated until it is judged at step ST
6
that the entropy coded data size is less than the set size.
In this repeated processing, the compression is performed while changing the quantization coefficient for each repetition, The quantization coefficient is selected and given to the quantization circuit
12
so that the compression rate becomes gradually higher.
When judging at step ST
6
that the entropy coded data size is less than the set size, a signal S
15
a
indicating the judgement is output to the JPEG file maker
17
by the quantization coefficient control circuit
15
.
Then, in the JPEG file maker
17
, a JPEG file based on the entropy coded compression image data D
13
of the size less than the set size is made and output (ST
8
).
Summarizing the problem to be solved by the invention, in the above image compression controller, due in part to the fact that the compression rate ends up becoming higher than necessary depending on the input image data, when trying to flexibly change the quantization coefficient for every image in order to obtain a better image quality within the limited capacity, the processing process increases due to the repetition of compression as explained above etc. and the processing time ends up increasing as well along with the number of repetitions.
Further, since the number of repetitions differs in accordance with an input image, the processing time is liable to end up becoming uneven.
Especially, entropy coding generally requires a long processing time, so repetition of the processing many times has a large effect on the overall processing time.
Also, even if increasing the clock frequency or tinkering with the circuit configuration so as to increase the speed of processing of each block, the processing is wasted the more the number of repetitions and is difficult to make the best use of the improved performance.
SUMMARY OF THE INVENTION
An object of the present invention is to provide an image compression controller capable of simplifying the quantization coefficient control in an image compression process and increasing the speed of image compression and an image pickup system using the same.
According to a first aspect of the present invention, there is provided an image compression controller for compressing image data to generate an image file having a size smaller than a set size, comprising a compression circuit for compressing redundancy of input image data for coding; a quantization circuit for first quantization based on a first quantization coefficient and second quantization based on a second quantization coefficient on an output signal of the compression circuit; a counter for counting the number of a quantization value having correlation with a size of an image file generated among quantization values included at least in first quantization data from the quantization circuit; a quantization coefficient control circuit for generating a control coefficient based on the count of the counter and a reference value of a quantization value having correlation with a size of an image file generated set in advance; a quantization coefficient generation circuit for supplying a reference quantization coefficient set in advance as the first quantization coefficient to the quantization circuit, generating the second quantization coefficient based on the control coefficient and the reference quantization coefficient, and supplying the same to the quantization circuit when a control coefficient from the quantization coefficient control circuit is received; and an entropy coding circuit for entro
Inagaki Takunari
Shimura Katsuharu
Couso Jose L.
Frommer William S.
Frommer & Lawrence & Haug LLP
Sony Corporation
LandOfFree
Image compression controller and image pickup system using... 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 controller and image pickup system using..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image compression controller and image pickup system using... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3072533