Image analysis – Image compression or coding – Predictive coding
Reexamination Certificate
2000-03-01
2003-10-21
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Image compression or coding
Predictive coding
C382S239000, C382S245000, C375S240140, C348S411100, C341S063000, C358S426130
Reexamination Certificate
active
06636642
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to technology for compressing image data, particularly relates to lossless compression for a multivalued input image.
2. Description of the Related Art
As the quantity of image data is generally enormous, image data is often compressed to reduce the quantity in communication, storage and others. A method of compressing image data is roughly classified into two types of a lossless coding method and a lossy coding method. In the latter case, visual degradation may occur in a decoded image depending upon compression ratio.
Recently, for a typical image described in page description language (PDL), an image generated by a computer (hereinafter generically called a computer graphic (CG) image) increases. As these images include no noise component caused in scanning, the degradation of image quality by lossy coding is remarkable at an edge.
In lossy coding called a baseline process defined in a joint photographic coding experts group (JPEG) (refer to p. 160 to 182 of “Interface”, December 1991 issue) which is a standard of image coding, after an image is converted to a frequency component using discrete cosine transform (DCT), it is quantized. JPEG is effective in a scanned image the energy of which concentrates on a low-frequency component. However, the degradation occurs in an image acquired by decoding a CG image including many high-frequency components. For an example of the degradation of image quality, mosquito noise caused in the vicinity of an edge and block noise caused between blocks can be given.
In the meantime, lossless coding, in which the same decoded image as an input image is acquired, can avoid the degradation of image quality. For an example of technology particularly related to a CG image in conventional type lossless coding, proposal disclosed in Japanese published unexamined patent application No. Hei 9-224253 will be described below. The above conventional type utilizes the characteristics of a CG image that a prediction based upon neighbor pixels often hit and uses a hit one of plural predictions as a code.
Conventional type image coding device and image decoding device are respectively shown in upper and lower halves in FIG.
16
. As shown in
FIG. 16
, a reference number
10
denotes an image input unit,
20
denotes a first predictor,
21
denotes a second predictor,
30
denotes a prediction error calculator,
31
denotes a prediction error adder,
40
and
41
denote a selector,
50
denotes an encoder,
51
denotes a decoder,
60
denotes a code output unit,
70
denotes a code input unit,
80
denotes an image output unit,
100
denotes image data,
110
denotes prediction value data,
111
denotes control data,
120
denotes prediction error data,
130
denotes prediction result data and
140
denotes code data.
Each unit shown in
FIG. 16
will be described below. The image coding device shown in the upper half in
FIG. 16
is made up as follows. The image input unit
10
receives an image input from an external device and respectively sends it to the first and second predictors
20
and
21
and the prediction error calculator
30
as image data
100
. The first and second predictors
20
and
21
predict the value of a target pixel based upon the image data
100
according to respective predetermined methods and send it to the selector
40
as predicted value data
110
. The prediction error calculator
30
predicts the value of the target pixel based upon the image data
100
according to a predetermined method, subtracts the predicted value from the actual value of the target pixel and sends the value to the selector
40
as prediction error data
120
. The selector
40
detects based upon the image data
100
and the predicted value data
110
whether the prediction is coincident at the target pixel or not. If there is one or more hit predictors, the selector converts the identification number of the hit predictor to prediction result data
130
and sends it to the encoder
50
. If not, the prediction error data
120
. The encoder
50
encodes the prediction result data
130
using a predetermined entropy coding method and sends it to the code output unit
60
as code data
140
. The code output unit
60
outputs the code data
140
to an external device as an output code.
Next, each unit of the image decoding device shown in the lower half in
FIG. 16
will be described. The same number is allocated to the same each unit as that in the image coding device and the description is omitted. The code input unit
70
receives a code input from an external device and sends it to the decoder
51
as code data
140
. The decoder
51
decodes the code data
140
according to a decoding method corresponding to the coding method used in the encoder
60
and sends it to the selector
41
as prediction result data
130
. The selector
41
sends control data
111
to the corresponding predictor if the prediction result data
130
represents the identification number of the predictor and instructs the predictor to output image data
100
. The selector also sends the prediction error data
120
to the prediction error adder
31
if the contents of the prediction result data
130
show an error. The prediction error adder
31
predicts the value of a target pixel according to the same method as that of the prediction error calculator
30
, adds the prediction error data
120
to the predicted value and sends it to the image output unit
80
as image data
100
. The image output unit
80
outputs the image data
100
to an external device to as an output image.
The operation based upon the above configuration will be described below.
FIGS. 17 and 18
are flowcharts showing the operation of conventional type image coding/decoding device. The operation of the conventional type will be described below.
First, referring to
FIG. 17
, coding processing will be described. In S
10
, the image input unit
10
inputs an image. In S
20
, the first and second predictors
20
and
21
predict the value of a target pixel based upon image data
100
and simultaneously, the prediction error calculator
30
calculates an error of a predicted value according to a predetermined method. In S
30
, the selector
40
sends the identification number of the hit predictor or a prediction error to the encoder
50
as prediction result data
130
. In S
50
, the encoder
50
applies predetermined coding to the prediction result data
130
. In S
60
, if input image data is finished, coding processing is terminated and if not, the processing proceeds to S
10
.
Next, decoding processing will be described referring to FIG.
18
. In S
110
, the code input unit
70
inputs a code. In S
120
, the decoder
51
executes decoding processing which is equivalent to processing reverse to coding processing executed by the encoder
50
and generates prediction result data
130
. In S
130
, the selector
41
sends control data
111
to either of the first or second predictor
20
or
21
having the following identification number when the prediction result data
130
represents the identification number of the predictor. If the prediction result data
130
represents a prediction error, the selector
41
also sends the prediction error to the prediction error adder
31
as prediction error data
120
. In S
21
, if there is one or more hit prediction, either of the first or second predictor
20
or
21
which receives the control data
111
executes predetermined prediction and generates image data
100
. If there is not, the prediction error adder
31
adds prediction error data
120
to a predicted value acquired by the similar prediction to the prediction error calculator
30
and generates image data
100
. In S
140
, the image output unit
80
outputs the image data
100
to an external device. In S
150
, when input code data is finished, decoding processing is terminated and if not, the processing proceeds to S
110
.
Of the above operation, the selection of the predictor in S
30
is performed according to a predeterm
Fuji 'Xerox Co., Ltd.
Mehta Bhavesh M.
Morgan & Lewis & Bockius, LLP
Sukhaphadhana Christopher
LandOfFree
Image coding device, image decoding device, 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 device, image decoding device, image..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image coding device, image decoding device, image... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3141112