Image coding apparatus and image coding method

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to an image coding apparatus and an image coding method for compressively coding image data.

2. Description of the Related Art

Image data generally falls into two broad categories: artificial image data generated by computer such as image data described in PDL (page description language), and scan-in image data read by scanner such as bit map image data. These two kinds of image data are most often coded compressively in order to reduce their quantities.

Illustratively, JPEG-LS (Part-1 ISO/IEC FDIS 14495-1) is known as a coding technique applicable to both artificial image data and scan-in image data. JPEG-LS has been standardized as a method for improving by about 10 percent the coding efficiency achieved by JPEG (Joint Photographic Expert Group) and JBIG (Joint Bi-level Image Experts Group) schemes. A major characteristic of the JPEG-LS method is that it applies lossless (reversible) coding to artificial image data and near-lossless (near-reversible) coding to scan-in image data.

FIG. 13

is a schematic block diagram of a typical image coding apparatus adopting JPEG-LS for image data coding. This image coding apparatus works as follows: input image data

401

and decoded pixel data

402

are first fed to a window composing element

41

. In turn, the window composing element

41

sends a group of peripheral pixel values

403

to a pixel value gradient determining element

42

. Given the peripheral pixel values, the pixel value gradient determining element

42

computes a pixel value gradient of the peripheral pixels surrounding a target pixel to be coded. If the computed gradient is found less than a tolerance

404

coming from a tolerance designating element

43

, the target pixel to be coded is judged flat and control is passed on to a run length extracting element

44

. If the gradient is found greater than the tolerance

404

, then regular mode is entered. That is, a pixel value predicting element

45

predicts a value of the target pixel to be coded on the basis of the peripheral pixel value group

403

. A prediction error computing element

46

computes a prediction error regarding the target pixel. A prediction error quantizing element

47

quantizes the computed prediction error. Taking the quantized prediction error as a coding symbol

406

, an information source coding element

48

codes the prediction error. The coded output of the element

48

combines with a code from the run length extracting element

44

to become output coded data

407

. A local decoding element

49

performs local decoding and supplies the result as decoded pixel data

402

to the window composing element

41

.

The JPEG-LS method outlined above is based on pixel value prediction and has no recourse to resolution adaptation known as sub-sampling. Sub-sampling involves discarding the whole information furnished by any one pixel. In low frequency image regions, particularly in graduated portions of scan-in image data, sub-sampling should preferably be utilized to enhance the efficiency of coding. In the setup of

FIG. 13

, however, the pixel value gradient determining element

42

determines whether or not to proceed with sub-sampling based on the gradient of pixels around the target pixel to be coded. In other words, the sub-sampling of the setup has no direct relation to prediction error.

Meanwhile, some of the image data coding techniques proposed so far have introduced the concept of sub-sampling into predicted coding. Illustratively, Japanese Published Unexamined Patent Application No. Hei 10-294872 discloses a sub-sampling-based coding technique for selecting an index (identification information) of one of multiple predictors that has the smallest prediction error so that a run length of the rank of the selected index is coded.

More particularly, as shown in

FIG. 14

, a group of peripheral pixel values

503

representing pixels around a target pixel to be coded is acquired through a window composing element

51

. Given the peripheral pixel value group

503

, a pixel value approximating element

52

having multiple predictors computes a group of decoded pixel value candidates

504

. An error selecting element

54

checks for a difference between each of the pixel values approximated by the pixel value approximating element

52

on the one hand, and the value of a target pixel in an image of interest on the other hand. If the difference (i.e., prediction error) is found less than a tolerance

505

coming from a tolerance designating element

53

, then the error selecting element

54

sends an index of the predictor having the smallest prediction error to a decoded pixel selecting element

55

and a rank/run length extracting element

56

. If the prediction error is judged to be greater than the tolerance

505

, the error selecting element

54

supplies the rank/run length extracting element

56

with an index of an error minimizing approximation

506

for minimizing the prediction error as well as with the value of the prediction error in effect at that point. Based on the error minimizing approximation index

506

or the prediction error value received from the error selecting element

54

, the rank/run length extracting element

56

extracts a run length of the index rank and sends it as a coding symbol

507

to an information source coding element

57

. The information source coding element

57

performs information source coding to compress the statistical redundancy included in the coding system

507

to provide output coded data

508

. The decoded pixel selecting element

55

, given the error minimizing approximation index

506

, selects a decoded pixel value from the group of decoded pixel value candidates

504

and feeds it as decoded pixel data

502

to the window composing element

51

for decoding of the next pixel.

In the manner described above, the coding technique of Japanese Published Unexamined Patent Application No. Hei 10-294872 adopting the concept of sub-sampling subjects artificial image data and scan-in image data to the same framework of reversible and near-reversible coding, and executes an image coding process by coding indexes of prediction techniques (predictors) for approximating each target pixel to be coded. That is, the approximation techniques (predictors) based on important factors of images are adopted for data compression without deterioration of image quality.

However, the conventional image data coding method disclosed by the above-cited patent application has its share of deficiencies. The method, in selecting one of the prediction techniques (predictors) for approximating the target pixel to be coded, abides by a single rule dictating that the selected technique be one that has the smallest prediction error. Hence the possible major flaw of the method as outlined below.

With pixel values making up artificial image data, efforts to reduce their prediction errors translate into improvements in image quality of non-reversible coding. Pixel values constituting scan-in image data, on the other hand, undergo MTF (modulation transfer function) illustratively in the optics of a scanner and otherwise gather various noises along the way. For that reason, minimizing the prediction error of each pixel does not necessarily result in improvements in image quality of non-reversible coding.

The above-cited conventional image data coding technique, particularly when addressing scan-in image data and when their prediction errors exceed a tolerance, is liable not only to worsen its coding efficiency but also to slow down its processing speed. That is because the prediction errors themselves are coded. Even if the prediction errors are below the tolerance, the selected predictor tends to change frequently as long as the sole judging rule dictates that the prediction error be minimal per pixel. The tendency defeats the supposed effects of coding run lengths of index ranks, rendering the coding process inefficient. Furthermore, improvements in image quality cannot be expected unless predictors are selected not on the basis of pixel-by-pixel local information but based on more extensive information.

SUMMARY OF THE INVENTION

The present invention has been made in view of the above circumstances and provides an image coding apparatus and an image coding method which, when subjecting artificial and scan-in image data to the same framework of reversible or near-reversible coding, select a technique for predicting a target pixel to be coded not on the basis of pixel-by-pixel local information alone but based on more extensive information, thereby boosting coding efficiency and enhancing image quality in non-reversible coding without recourse to bulky computation resources or massive calculation processes.

In carrying out the invention and according to one aspect thereof, there is provided an image coding apparatus including an approximating element which has plural different prediction techniques to approximate, in input image data, a value of a target pixel to be coded with reference to values of peripheral pixels surrounding the target pixel, and a holding element which holds information about ranks of the prediction techniques. The ranks are obtained upon processing of the pixel preceding the target pixel. The apparatus also includes a determining element which computes an error between each of approximate values obtained by the plural prediction techniques and the value of the target pixel, and which selects for the target pixel one of the prediction techniques based on the computed error. The determining element performs the selection by preferentially referencing the rank information held in the holding element if the error is within a predetermined tolerance. The apparatus further includes a coding element which codes the value of the target pixel using the approximate value obtained by the prediction technique selected by the determining element.

According to another aspect of the invention, there is provided an image coding apparatus including an approximating element which has plural different prediction techniques to approximate, in input image data, a value of a target pixel to be coded with reference to values of peripheral pixels surrounding the target pixel, and a holding element which holds information about run lengths approximated by the prediction techniques. The run lengths are obtained upon processing of each pixel preceding the target pixel. The apparatus also includes a determining element which computes an error between each of approximate values obtained by the plural prediction techniques and the value of the target pixel, and which selects for the target pixel one of the prediction techniques based on the computed error. The determining element performs the selection by preferentially referencing the run lengths specified by the information held in the holding element if the error is within a predetermined tolerance. The apparatus further includes a coding element which codes the value of the target pixel using the approximate value obtained by the prediction technique selected by the determining element.

According to a further aspect of the invention, there is provided an image coding method including the steps of approximating, in input image data, a value of a target pixel to be coded with reference to values of peripheral pixels surrounding the target pixel by using plural different prediction techniques, and computing an error between each of approximate values obtained by the prediction techniques and the value of the target pixel, to determine whether the computed error is within a predetermined tolerance. The method further includes the steps of, if the error is found to be within the predetermined tolerance and if the prediction techniques have been given ranks beforehand, selecting one of the prediction techniques for the target pixel by preferentially referencing the ranks, if the error is not within the predetermined tolerance and if the prediction techniques are not given ranks, then selecting one of the prediction techniques for addressing the target pixel on the basis of differences in magnitude between the errors regarding the approximate values, and coding the target pixel using the approximate value provided by the selected prediction technique.

According to an even further aspect of the invention, there is provided an image coding method including the steps of approximating, in input image data, a value of a target pixel to be coded with reference to values of peripheral pixels surrounding the target pixel by using plural different prediction techniques, and computing an error between each of approximate values obtained by the plural prediction techniques and the value of the target pixel, to determine whether the computed error is within a predetermined tolerance. The method further includes the steps of, if the error of any one of the approximate values obtained by any one of the prediction techniques is found to be within the predetermined tolerance and if the target pixel is shifted successively from one pixel position to another within the input image data, determining whether the errors of the approximate values obtained by the same prediction technique in different pixel positions fall successively within the predetermined tolerance, if the errors of pixels having run lengths are found to be successively within the predetermined tolerance, then selecting one of the prediction techniques that provides the largest run length as the prediction technique for a group of pixels corresponding to the largest run length, if the computed error is not found to be within the predetermined tolerance and if the errors in different pixel positions are not successively found to be within the tolerance, then selecting one of the prediction techniques for the target pixel on the basis of differences in magnitude between the errors regarding the approximate values, and coding the target pixel using the approximate value provided by the selected prediction technique.

From one point of view, the inventive image coding apparatus and image coding method determine, upon selecting one of multiple prediction techniques to approximate the value of target pixel to be coded, whether the error of each approximated value falls within a predetermined tolerance. If the error of the value approximated by each prediction technique is found to be within the tolerance and if the prediction techniques have been given ranks beforehand, these ranks are preferentially referenced in selecting one of the prediction techniques for addressing the target pixel. That is, not only the prediction error per pixel but also the rank of each prediction technique with respect to the preceding pixel is taken into account in selecting the appropriate prediction technique. It follows that as long as the errors fall within the tolerance, the ranks of the prediction techniques will not change frequently.

From another point of view, the inventive image coding apparatus and image coding method determine, upon selecting one of multiple prediction techniques to approximate the value of target pixel to be coded, whether the errors of the approximate values obtained by the same prediction technique in different pixel positions fall successively within the predetermined tolerance. If the errors are found to be successively within the predetermined tolerance, one of the prediction techniques is selected preferentially in terms of run lengths. That is, since not only the prediction errors per pixel but also their run lengths are used as a selection rule, what is preferentially selected from among the prediction techniques with their errors falling within the tolerance is the technique having the largest run length.

Other features and advantages of the invention will become more apparent upon a reading of the following description and appended drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Preferred embodiments of the present invention will be described in detail with reference to the following figures wherein:

FIG. 1

is a schematic block diagram of an image coding apparatus practiced as part of a first embodiment of this invention;

FIG. 2

is a flowchart of steps constituting an image coding method implemented as part of the first embodiment, the flowchart showing more particularly steps performed by a predictor determination circuit in the image coding apparatus of

FIG. 1

;

FIG. 3

is a flowchart of steps constituting the image coding method implemented as part of the first embodiment, the flowchart showing more particularly steps performed by a preceding pixel index holding circuit in the image coding apparatus of

FIG. 1

;

FIG. 4

is a schematic explanatory view that helps explain prediction techniques used by the image coding method according to the invention;

FIG. 5

is an explanatory view outlining prediction techniques used by the inventive image coding method;

FIG. 6

is a schematic block diagram of an image coding apparatus practiced as part of a second embodiment of this invention;

FIG. 7

is a flowchart of steps constituting an image coding method implemented as part of the second embodiment, the flowchart showing more particularly steps performed by a predictor determination circuit in the image coding apparatus of

FIG. 6

;

FIG. 8

is a flowchart of steps constituting the image coding method implemented as part of the second embodiment, the flowchart showing more particularly steps performed by a total predictor rank/run length extraction circuit in the image coding apparatus of

FIG. 6

;

FIG. 9

is a flowchart of steps constituting the image coding method implemented as part of the second embodiment, the flowchart showing more particularly steps performed by a total predictor rank/run length holding circuit in the image coding apparatus of

FIG. 6

;

FIG. 10

is a schematic block diagram of an image coding apparatus practiced as part of a third embodiment of this invention;

FIG. 11

is a flowchart of steps constituting an image coding method implemented as part of the third embodiment, the flowchart showing more particularly steps performed by a predictor determination circuit in the image coding apparatus of

FIG. 10

;

FIG. 12

is a flowchart of steps constituting the image coding method implemented as part of the third embodiment, the flowchart showing more particularly steps performed by a cumulative prediction error holding circuit in the image coding apparatus of

FIG. 10

;

FIG. 13

is a schematic block diagram of an image coding apparatus based on the conventional JPEG-LS coding method; and

FIG. 14

is a schematic block diagram of an image coding apparatus based on another conventional coding method.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

First Embodiment

An image coding apparatus and an image coding method practiced as the first embodiment are described below. The image coding apparatus is first described by referring to

FIG. 1

, i.e., a schematic block diagram of the apparatus.

As shown in

FIG. 1

, the image coding apparatus of the first embodiment is constituted by a window composition circuit

11

, multiple predictors

12

a

through

12

e

, a predictor determination circuit

13

, a tolerance designation circuit

14

, a rank updating circuit

15

, a rank/run length extraction circuit

16

, a Huffman coding circuit

17

, a preceding pixel index holding circuit

18

, and a decoded pixel selection circuit

19

. Each of the component circuits is described below in detail.

The window composition circuit

11

receives input image data

101

that are entered illustratively pixel by pixel in raster scan sequence. From the input image data

101

, the window composition circuit

11

extracts a value of a target pixel to be coded (i.e., pixel of interest), as well as decoded values of a group of peripheral pixels surrounding the target pixel. The decoded values thus extracted are sent as a peripheral pixel value group

102

to the predictors

12

a

through

12

e

and to the predictor determination circuit

13

. The values of the peripheral pixels around the target pixel are extracted on the basis of decoded pixel data

103

coming from the decoded pixel selection circuit

19

as will be described later.

Given the peripheral pixel value group

102

from the window composition circuit

11

, each of the predictors

12

a

through

12

e

approximates through prediction a value of the target pixel using a different prediction technique. The approximated values are sent as a decoded pixel value candidate group

104

to the predictor determination circuit

13

and decoded pixel selection circuit

19

. In other words, the predictors

12

a

through

12

e

function as the approximating element of this invention. It is assumed that prediction techniques of the predictors

12

a

through

12

e

each include multiple quantization and sub-sampling (interpolation) techniques utilizing a group of decoded pixel values. Illustratively, one quantization technique may involve using for quantization the largest of a group of decoded pixel values on the basis of: an average of the pixels immediately left and immediately on top of a pixel of interest; an average of a group of all peripheral pixels surrounding the pixel of interest; a mean value greater than the average of the group of all peripheral pixels surrounding the pixel of interest; a mean value less than the average of the group of all peripheral pixels surrounding the pixel of interest; or prediction errors themselves. Another quantization technique may involve resorting to medians, and another quantization technique may be an interpolation technique utilizing an average of a group of decoded pixel values as an interpolated value. All these techniques are designed to approximate the value of the target pixel to be coded.

The predictor determination circuit

13

computes errors in terms of differences between the value of the pixel of interest included in a peripheral pixel value group

102

from the window composition circuit

11

on the one hand, and a decoded pixel value candidate group

104

from the predictors

12

a

through

12

e

, i.e., approximate values of the pixel of interest predicted by the predictors

12

a

through

12

e

on the other hand. The prediction errors thus obtained are compared with a tolerance

105

designated by the tolerance designation circuit

14

. As a result of the comparison, the predictor determination circuit

13

selects one of the predictors for addressing the pixel of interest. The result of the selection is sent to the rank updating circuit

15

, rank/run length extraction circuit

16

, and decoded pixel selection circuit

19

as a global pixel information priority technique index

106

. If the errors of prediction by all predictors

12

a

through

12

e

are found to exceed the tolerance

105

, the predictor determination circuit

13

selects preferentially the predictor with the smallest prediction error. If some of the prediction errors are found to be within the tolerance

105

, then the predictor determination circuit

13

selects from among the predictors associated with the tolerance-compliant errors a predictor which meets a specific condition. That condition dictates that if the preceding pixel index holding circuit

18

, to be described later, has in its contents the same ranking index as that adopted for the preceding pixel, the predictor applicable to that index should be selected unconditionally; otherwise the predictor whose prediction error is the smallest of those within the tolerance

105

should be selected. That is, the predictor determination circuit

13

functions as the determining element of this invention.

The tolerance designation circuit

14

retains a predetermined tolerance

105

. At the same time, the circuit

14

designates the tolerance

105

to the predictor determination circuit

13

.

Based on the global pixel information priority technique index

106

from the predictor determination circuit

13

, the rank updating circuit

15

determines information

107

on correspondence between approximations and ranks regarding the predictors

12

a

through

12

e

. The information

107

is forwarded to the rank/run length extraction circuit

16

. The rank updating circuit

15

updates ranks according to the global pixel information priority technique index

106

after any one target pixel has been coded. The updating of ranks may be done in order of certainty. For example, the most recently selected approximation technique may be ranked first; the most frequently selected approximation technique in the coding of several past pixels may be ranked first; or the first rank may be accorded to the approximation technique judged to have the largest run length as a result of the coding of several past pixels.

Given the approximation-and-rank correspondence information

107

from the rank updating circuit

15

, the rank/run length extraction circuit

16

determines a rank relevant to the global pixel information priority technique index

106

in counting up a rank run length. The rank run length is sent as a coding symbol

108

to the Huffman coding circuit

17

and preceding pixel index holding circuit

18

. If first ranks are found to continue, the rank/run length extraction circuit

16

outputs their continuous length, i.e., run length. If a prediction error received as a global pixel information priority technique index

106

is found to exceed the tolerance, the rank/run length extraction circuit

16

passes on the error, unmodified, as the coding symbol

108

for output.

The Huffman coding circuit

17

receives the coding symbol

108

from the rank/run length extraction circuit

16

, codes it into an information source, and sends the coded result as output coded data

109

to the outside of the image coding apparatus. More specifically, the Huffman coding circuit

17

refers to the Huffman coding table when coding the coding symbol

108

. The Huffman coding scheme for information source coding may be replaced by some other suitable scheme for suppressing statistical redundancy in coding symbols, such as arithmetic coding. The Huffman coding circuit

17

functions as the coding element of this invention.

The preceding pixel index holding circuit

18

receives and temporarily stores the coding symbol

108

from the rank/run length extraction circuit

16

. The received symbol is then sent to the predictor determination circuit

13

as a ranking index (called the preceding pixel index)

110

adopted for the pixel preceding the current pixel of interest. That is, the preceding pixel index holding circuit

18

functions as the holding element of this invention.

The decoded pixel selection circuit

19

selects from among the group of decoded pixel value candidates

104

a candidate which corresponds to the global pixel information priority technique index

106

. The selected candidate is output as decoded pixel data

103

for use by the window composition circuit

11

in extracting peripheral pixel values.

How the image coding apparatus of the above-described constitution illustratively works will now be described.

FIGS. 2 and 3

are flowcharts of steps constituting the image coding method as part of the first embodiment of this invention.

FIGS. 4 and 5

are explanatory views outlining typical prediction techniques used by the image coding method.

As shown in

FIG. 4

, the window composition circuit

11

of the image coding apparatus initially receives input image data

101

on a pixel-by-pixel basis in raster scan sequence, and regards each pixel as a target pixel X to be coded. The target pixel X and previously decoded peripheral pixels R

0

through R

11

around the target pixel are sent collectively to the predictors

12

a

through

12

e

as a peripheral pixel value group

102

. The predictors

12

a

through

12

e

each approximate the target pixel X using different prediction techniques (H, L, T, B, E . . . ), as shown in FIG.

5

.

After the predictors

12

a

through

12

e

have approximated values of the target pixel X, the predictor determination circuit

13

of the image coding apparatus compares the approximate values of the target pixel from the predictors with the target pixel value included in the peripheral pixel value group

102

from the window composition circuit

11

, as indicated in FIG.

2

. The comparison yields prediction errors of the predictors

12

a

through

12

e

(in step

1001

; hereinafter, step is referred to as “S”). The predictor determination circuit

13

checks to see if each of the computed prediction errors falls within the tolerance

105

designated by the tolerance designation circuit

14

(S

1002

).

If none of the prediction errors is found to be within the tolerance

105

, the predictor determination circuit

13

selects from among the predictors

12

a

through

12

e

the predictor that has the smallest of errors in magnitude. The prediction error associated with the selected predictor is output as a global pixel information priority technique index

106

(S

1003

).

If any one of the prediction errors is judged to be within the tolerance

105

, the predictor determination circuit

13

checks to see whether the predictor associated with the tolerance-compliant error has the same preceding pixel index

110

as that held by the preceding pixel index holding circuit

18

, i.e., whether the predictor in question is the same as the predictor which was identified by the preceding pixel index

110

and which was selected in processing of the pixel preceding the target pixel (S

1004

).

If the predictor in question has the same index as the preceding pixel index

110

, the predictor determination circuit

13

unconditionally adopts the preceding pixel index

110

and outputs it as the global pixel information priority technique index

106

(S

1006

). If the predictor in question is found to have an index different from the preceding pixel index

110

, the predictor determination circuit

13

selects the predictor that has yielded the smallest of the prediction errors complying with the tolerance

105

. The prediction error associated with the selected predictor is then output as the global pixel information priority technique index

106

(S

1005

).

That is, when determining the global pixel information priority technique index

106

, the predictor determination circuit

13

selects the predictor that has the smallest of prediction errors as long as the prediction errors of all predictors

12

a

through

12

e

exceed the tolerance

105

. If some of the prediction errors fall within the tolerance

105

, the predictor determination circuit

13

selects either the predictor identified by the preferentially referenced contents of the preceding pixel index

110

or the predictor that has the smallest prediction error. In this manner, a predictor meeting a specific condition is selected from among the predictors whose prediction errors have been computed.

The predictor determination circuit

13

checks to see whether all pixels making up the input image data

101

have been processed (S

1007

). The steps above are repeated until all pixels have been dealt with (S

1001

through S

1007

).

Given the global pixel information priority technique index

106

from the predictor determination circuit

13

, the rank/run length extraction circuit

16

outputs either the coding symbol

108

or the run length accordingly. The Huffman coding circuit

17

codes the coding symbol

108

by referring illustratively to the Huffman coding table. The coded result is output as the output coded data

109

.

At this point, the preceding pixel index holding circuit

18

holds the preceding pixel index

110

by following the steps shown in FIG.

3

. The preceding pixel index holding circuit

18

first checks to see whether the coding symbol

108

from the rank/run length extraction circuit

16

, i.e., the global pixel information priority technique index

106

from the predictor determination circuit

13

, is a prediction error (S

2001

). In other words, the preceding pixel index holding circuit

18

checks to see if the global pixel information priority technique index

106

has been determined on the basis of differences in magnitude between the prediction errors of the predictors

12

a

through

12

e.

If the global pixel information priority technique index

106

is not found to be a prediction error, the preceding pixel index holding circuit

18

judges that there was a prediction error complying with the tolerance

105

upon determination of the index

106

by the predictor determination circuit

13

and that the index

106

was the same as the preceding pixel index

110

. The global pixel information priority technique index

106

is then preserved as the preceding pixel index

110

(S

2002

). The preceding pixel index

110

thus retained is made available for use by the predictor determination circuit

13

in processing the next target pixel to be coded.

If the global pixel information priority technique index

106

is found to be a prediction error, then the preceding pixel index holding circuit

18

judges that the continuity of equally ranked indexes has been broken and proceeds to erase the retained preceding pixel index

110

(S

2003

). At this point, the preceding pixel index holding circuit

18

substitutes the prediction error in question for the erased preceding pixel index

110

and retains the error. The prediction error thus held is made available for use by the predictor determination circuit

13

in processing the next target pixel to be coded.

The preceding pixel index holding circuit

18

checks to see if all pixels making up the input image data

101

have been processed (S

2004

). The steps above are repeated until all pixels have been dealt with (S

2001

through S

2004

).

According to the image coding apparatus and image coding method of the first embodiment, as described above, one of the multiple predictors

12

a

through

12

e

is selected to approximate the value of the target pixel to be coded. In that case, if the errors of prediction by the predictors

12

a

through

12

e

fall within the tolerance

105

and if the preceding pixel index

110

is being held, then the retained index

110

is applied preferentially to the current target pixel as well. Thus as long as the prediction errors comply with the tolerance

105

, the ranking indexes of the predictors

12

a

through

12

e

do not change frequently.

That is, the image coding apparatus or image coding method of the first embodiment adopts as a selection rule not only the prediction errors in increments of pixels but also the ranking indexes of the predictors

12

a

through

12

e

in effect in processing of the preceding pixel. This prevents the ranking indexes of the predictors

12

a

through

12

e

from getting altered frequently and thereby enhances coding efficiency while forestalling deterioration in image quality due to tolerance restrictions. Concurrently the burdens of coding computations are alleviated. Where not only artificial image data but also noise-plagued scan-in image data are to be dealt with, the ranking indexes of the predictors are turned into run lengths through prediction coding by preferentially regarding the ranking index of the predictor that determined the preceding pixel value as representative of a more extensive area of pixels, whereby coding efficiency is improved.

Second Embodiment

An image coding apparatus and an image coding method practiced as the second embodiment are described below. A constitution of the image coding apparatus is described first.

FIG. 6

is a schematic block diagram of the image coding apparatus practiced as the second embodiment of this invention. In

FIG. 6

, the component parts having the same or corresponding functions as their counterparts of the first embodiment are designated by like reference numerals, and their descriptions are omitted as redundant. The description below highlights differences between the two embodiments.

As illustrated, what makes the image coding apparatus of the second embodiment specifically different from the first embodiment is the presence of a predictor determination circuit

21

, a total predictor rank/run length extraction circuit

22

, and a total predictor rank run length holding circuit

23

. Each of these circuits is described below in detail.

The predictor determination circuit

21

receives: a peripheral pixel value group

102

from the window composition circuit

11

, approximate values of a target pixel predicted by the predictors

12

a

through

12

e

, and a tolerance

105

from the tolerance designation circuit

14

. In turn, the predictor determination circuit

21

selects one of the predictors for processing the target pixel, and sends the result of the selection to the rank updating circuit

15

, decoded pixel selection circuit

19

and total predictor rank/run length extraction circuit

22

as a global pixel information priority technique index

201

. It should be noted that the predictor determination circuit

21

regards maximum run length information

203

from the total predictor rank run length holding circuit

23

(to be described later) as a determining rule for predictor selection in disregard of the preceding pixel index

110

used by the first embodiment. That is, the predictor determination circuit

21

functions as the determining element of this invention.

The total predictor rank/run length extraction circuit

22

receives the global pixel information priority technique index

201

from the predictor determination circuit

21

and approximation-and-rank correspondence information

107

from the rank updating circuit

15

. In response, the circuit

22

yields a coding symbol

108

and sends it to the Huffman coding circuit

17

. Given the approximation-and-rank correspondence information

107

, the total predictor rank/run length extraction circuit

22

counts a continuous length (i.e., run length) regarding all predictors

12

a

through

12

e

and sends the result to the total predictor rank run length holding circuit

23

as total predictor run length information

202

.

The total predictor rank run length holding circuit

23

receives the total predictor run length information

202

from the total predictor rank/run length extraction circuit

22

for temporary retention. Based on the information thus retained, the circuit

23

yields run length information about all predictors

12

a

through

12

e

and sends it as maximum run length information

203

to the predictor determination circuit

21

. That is, the total predictor rank run length holding circuit

23

functions as the holding element of this invention holding the total predictor rank run length.

What follows is a description of the image coding method of the second embodiment, i.e., of how the image coding apparatus of the above-described constitution typically works.

FIGS. 7

,

8

and

9

are flowcharts of steps outlining the image coding method of the second embodiment.

The predictors

12

a

through

12

e

in the image coding apparatus of the second embodiment, as in the first embodiment, approximate values of the target pixel X to be coded. As indicated in

FIG. 7

, the predictor determination circuit

21

computes a prediction error of each of the predictors (S

3001

). A check is made to see if each prediction error falls within the tolerance

105

(S

3002

).

If some of the prediction errors are judged to fall within the tolerance

105

, the predictor determination circuit

21

selects all predictors that have yielded the tolerance-compliant prediction errors. The result of the selection (i.e., index of each selected predictor) is output as a global pixel information priority technique index

106

(S

3003

).

If none of the prediction errors is found to be within the tolerance

105

, the predictor determination circuit

21

judges that the continuity of prediction errors complying with the tolerance

105

has been broken. In that case, based on the maximum run length information

203

from the total predictor rank run length holding circuit

23

, the predictor determination circuit

21

checks to see whether the number of runs before the interruption (i.e., continuous pixel count) is larger than one (S

3004

).

If the run count is larger than one, the predictor determination circuit

21

selects the predictor having a longer run length than any other predictor on the basis of the maximum run length information

203

. The result of the selection (i.e., index of the selected predictor) is output as the global pixel information priority technique index

106

(S

3005

). If the run count is not in excess of one, the predictor determination circuit

21

recognizes a run discontinuity and thus selects the predictor with the smallest prediction error. The prediction error of that predictor is output as the global pixel information priority technique index

106

(S

3006

).

That is, when determining the global pixel information priority technique index

106

, the predictor determination circuit

21

works as follows: if the prediction errors of the predictors

12

a

through

12

e

fall within the tolerance

105

, a check is made to see, as the target pixel is shifted from one pixel position to another within the input image data

101

, whether the prediction errors of the same predictor among those

12

a

through

12

e

fall continuously within the tolerance

105

. If the errors are found to comply continuously with the tolerance

105

, then the predictor having the largest run length is selected to address a group of pixels corresponding to the run length in question regardless of the ranks accorded to the predictors

12

a

through

12

e.

The predictor determination circuit

21

checks to see whether all pixels making up the input image data

101

have been processed (S

3007

). The steps above are repeated until all pixels have been dealt with (S

3001

through S

3007

).

Given the global pixel information priority technique index

106

from the predictor determination circuit

21

, the total predictor rank/run length extraction circuit

22

outputs a coding symbol

108

based on the index

106

. At this point, the total predictor rank/run length extraction circuit

22

outputs the coding symbol

108

using the steps shown in FIG.

8

.

Upon receipt of the global pixel information priority technique index

106

, the total predictor rank/run length extraction circuit

22

checks to see whether the index

106

is a prediction error (S

4001

). If the index

106

is not found to be a prediction error, the circuit

22

judges that the continuity of prediction errors complying with the tolerance

105

has not been broken. In that case, the total predictor rank/run length extraction circuit

22

supplies the total predictor rank run length holding circuit

23

with total predictor run length information

202

(S

4002

). The information

202

specifies that the received global pixel information priority technique index

106

be incremented, i.e., that the indexes of all predictors having yielded prediction errors in compliance with the tolerance

105

be added up.

If the global pixel information priority technique index

106

is not found to be a prediction error, the total predictor rank/run length extraction circuit

22

judges that the continuity of prediction errors complying with the tolerance

105

has been broken. The circuit

22

then checks to see if the number of runs before the interruption is greater than one (S

4003

). If the run count is found to be greater than one, the total predictor rank/run length extraction circuit

22

does not send the total predictor run length information

202

to the total predictor rank run length holding circuit

23

. Instead, the circuit

22

supplies the Huffman coding circuit

17

with a coding symbol

108

representing the global pixel information priority technique index

106

from the predictor determination circuit

21

, i.e., the index of the predictor selected as having the largest run length before the interruption of the prediction error continuity (S

4004

).

If the run count up to the error discontinuity is not greater than one, then the total predictor rank/run length extraction circuit

22

recognizes a complete absence of the continuity of prediction errors in compliance with the tolerance

105

. In that case, as with the first embodiment, the smallest prediction error received from the predictor determination circuit

21

is forwarded as a coding symbol

108

to the Huffman coding circuit

17

(S

4005

). At this point, too, the total predictor rank/run length extraction circuit

22

does not send the total predictor run length information

202

to the total predictor rank run length holding circuit

23

.

In the above state, the total predictor rank run length holding circuit

23

retains the total predictor run length information

202

using the steps shown in FIG.

9

. The circuit

23

first checks to see if the total predictor run length information

202

has been sent from the total predictor rank/run length extraction circuit

22

(S

5001

). This check is made every time a new target pixel to be coded is taken up.

If the total predictor run length information

202

is found being sent in, the total predictor rank run length holding circuit

23

receives and retains the information

202

(S

5002

). The run length information updated by use of the received total predictor run length information

202

(e.g., following an increment of the continuous run count) is sent as maximum run length information

203

to the predictor determination circuit

21

(S

5003

). If the total predictor run length information

202

is not found to be sent in, the total predictor rank run length holding circuit

23

recognizes a prediction error discontinuity and erases the retained total predictor run length information

202

accordingly (S

5004

).

The total predictor rank run length holding circuit

23

then checks to see if all pixels making up the input image data

101

have been processed (S

5005

). The steps above are repeated until all pixels have been dealt with (S

5001

through S

5005

).

According to the image coding apparatus and image coding method of the second embodiment, as described above, one of the predictors

12

a

through

12

e

is selected to approximate a target pixel to be coded. A check is made to see, as the target pixel is shifted from one pixel position to another, whether the prediction errors of the same predictor among those

12

a

through

12

e

fall continuously within the tolerance

105

. If the errors are found to comply continuously with the tolerance

105

, then the predictor having the largest run length is selected preferentially to address a group of pixels corresponding to the run length in question. That is, as long as the prediction errors of the predictors comply with the tolerance

105

, the predictor with the largest run length is selected preferentially regardless of differences in magnitude between the errors involved.

The image coding apparatus or image coding method of the second embodiment adopts as a selection rule not only the prediction errors in increments of pixels but also run lengths of the errors. Any of the predictors

12

a

through

12

e

is selected in such a manner that the largest run length remains in effect. This prevents deterioration in image quality due to predictor tolerance restrictions while enhancing coding efficiency. Concurrently the burdens of coding computations are alleviated. Where not only artificial image data but also noise-plagued scan-in image data are to be dealt with, the ranking indexes of the predictors are turned into run lengths through prediction coding involving the preferential use of the predictor with the largest run length. This promotes sub-sampling of the target pixel to be coded and further boosts coding efficiency.

Third Embodiment

An image coding apparatus and an image coding method practiced as the third embodiment are described below. A constitution of the image coding apparatus is described first.

FIG. 10

is a schematic block diagram of the image coding apparatus practiced as the third embodiment of this invention. As with the second embodiment, the component parts in

FIG. 10

having the same or corresponding functions as their counterparts in the first embodiment are designated by like reference numerals, and their descriptions are omitted as redundant. The description below focuses on differences between the third embodiment and the first or the second embodiment.

As illustrated in

FIG. 10

, the image coding apparatus of the third embodiment is a variation of the first embodiment supplemented with a cumulative prediction error holding circuit

31

.

The cumulative prediction error holding circuit

31

receives and accumulates errors computed by a predictor determination circuit

32

regarding each of the predictors

12

a

through

12

e

. In response to a request from the predictor determination circuit

32

, the cumulative prediction error holding circuit

31

sends the retained cumulative value to the circuit

32

as a cumulative prediction error value

301

. The cumulative prediction error holding circuit

31

functions as the cumulative error holding element of this invention.

The predictor determination circuit

32

of the third embodiment, compared with its counterpart in the first embodiment, has an additional function: in selecting one of the predictors

12

a

through

12

e

to address the target pixel, the predictor determination circuit

32

adds the cumulative error coming from the cumulative prediction error holding circuit

31

to the error computed with respect to the pixel in question. The sum of the errors is used as a basis for the circuit

32

to select the suitable predictor.

Below is a description of the image coding method of the third embodiment, i.e., of how the image coding apparatus of the above-described constitution typically works.

FIGS. 11 and 12

are flowcharts of steps outlining the image coding method of the third embodiment.

As with the first embodiment, the predictors

12

a

through

12

e

in the image coding apparatus of the third embodiment approximate values of a target value X to be coded. With the approximate values thus predicted, the predictor determination circuit

32

computes a prediction error of each of the approximated values as shown in

FIG. 11

(S

6001

).

With the prediction error computed, the predictor determination circuit

32

receives the cumulative prediction error value

301

about each of the predictors

12

a

through

12

e

from the cumulative prediction error holding circuit

31

, and adds the received value to the prediction error for each predictor (S

6002

). The predictor determination circuit

32

then checks to see if the sum of the errors falls within the tolerance

105

designated by the tolerance designation circuit

14

(S

6003

). The subsequent steps are the same as those of the first embodiment (S

6004

through S

6008

).

The predictor determination circuit

32

also supplies the cumulative prediction error holding circuit

31

with the results of the prediction error computation (i.e., after addition of the cumulative prediction error value

301

) regarding all predictors

12

a

through

12

e.

The cumulative prediction error holding circuit

31

accumulates the sum of the prediction errors and outputs the cumulative prediction error value

301

using the steps shown in FIG.

12

. The circuit

31

first receives (S

7001

) the result of the prediction error computation (i.e., after addition of the cumulative prediction error value

301

), and retains the received result as the cumulative prediction error value

301

(S

7002

). Upon receipt of a command or like request from the predictor determination circuit

32

requesting the cumulative prediction error value

301

(S

7003

), the cumulative prediction error holding circuit

31

returns to the circuit

32

(S

7004

) the cumulative prediction error value

301

which is in effect at that point and which has been accumulated up to the pixel preceding the current target pixel of interest to be processed by the circuit

32

.

The cumulative prediction error holding circuit

31

then checks to see if all pixels making up the input image data

101

have been processed (S

7005

). The steps above are repeated until all pixels have been dealt with (S

7001

through S

7005

).

According to the image coding apparatus and image coding method of the third embodiment, as described above, one of the predictors

12

a

through

12

e

is selected to approximate the target pixel to be coded. The cumulative prediction error value

301

regarding each of the predictors

12

a

through

12

e

is added up to provide a basis for predictor selection. That is, not only localized information about the peripheral pixels surrounding each target pixel but also more extensive information about past pixels is used as a determining rule by which to select the appropriate predictor.

By utilizing not only the localized information but also global information about pixels as a yardstick for predictor selection, the image coding apparatus or image coding method of the third embodiment offsets widespread prediction errors spanning the image pixel data

101

. If localized prediction errors are apparently small but amount to a significant prediction error when accumulated for the entire input image data

101

, it is possible to bypass the predictor that has yielded the error in its approximation. As with the first embodiment, the third embodiment promotes sub-sampling of the target pixel to be coded and enhances the image quality for more extensive areas while offering better coding efficiency than before.

Although the third embodiment has been shown selecting any of the predictors

12

a

through

12

e

in the same manner as the first embodiment, this is not limitative of the invention. Alternatively, predictors may be selected in the same manner as with the second embodiment with no less image quality enhancement.

To sum up, in submitting artificial and scan-in image data to the same framework of reversible or near-reversible coding, the inventive image coding apparatus and image coding method select the technique for predicting the target pixel to be coded not only on the basis of localized pixel information but also with reference to more extensive pixel value information which may illustratively cover previously assigned ranks, run lengths, or cumulative errors of the prediction techniques involved. Such features boost coding efficiency and enhance image quality in non-reversible coding without recourse to bulky computation resources or massive calculation processes (e.g., the loads of computations are one order of magnitude less than JPEG).

It is to be understood that while the invention has been described in conjunction with specific embodiments, it is evident that many alternatives, modifications and variations will become apparent to those skilled in the art in light of the foregoing description. Accordingly, it is intended that the present invention embrace all such alternatives, modifications and variations as fall within the spirit and scope of the appended claims.

The entire disclosure of Japanese Patent Application No. 2000-70763 filed on Mar. 14, 2000 including specification, claims, drawings and summary is incorporated herein by reference in its entirety.

Image coding apparatus and image coding method

Information

Patent Number

Date Filed

Date Issued

Inventors

Original Assignees

Examiners

Agents

CPC

US Classifications

Field of Search

US

International Classifications

Term Extension

Abstract

Description

Claims

Priority Claims (1)

US Referenced Citations (1)

Foreign Referenced Citations (1)

Non-Patent Literature Citations (1)