JPS62247688A - Orthogonal conversion coding method for picture data - Google Patents

Abstract

PURPOSE:To improve a data compression rate by adding successively each absolute value of conversion data toward the low frequency component from the high frequency component and zone sampling only the conversion data having a frequency lower than the conversion data obtained when said added absolute value reaches a prescribed level. CONSTITUTION:The picture data (X) showing a halftone picture undergoes the orthogonal conversion via an orthogonal conversion circuit 11 after a preprocessing via a preprocessing circuit 10. This conversion data (y) is outputted in a sequence based on the order of space frequencies and applied to a coding circuit 12. A bit distribution decoding part 20 is connected to the circuit 12 and receives the data (y) to add successively the absolute values of the data (y) toward the low frequency component from the high frequency component. Then a code table of >=1 bit is allocated to the data (y) only when said added absolute value reaches a prescribed level at a point where said data (y) exists. While bit 0 is allocated to the data (y) of the relevant value. The circuit 12 performs the coding operation according to the bit distribution information.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of the Invention) The present invention relates to a method of encoding image data for the purpose of data compression, and more particularly to a method of encoding image data using orthogonal transformation.

(Technical Background and Prior Art of the Invention) Image signals carrying halftone images, such as 1'V No. 48, have an enormous amount of information, so a broadband transmission line is required for its transmission. be. Conventionally, such image signals have a large amount of redundancy [1], and various attempts have been made to compress image data by suppressing this redundancy. Furthermore, recently, it has become common practice to record halftone images on optical disks, magnetic disks, etc., and in this case, image data compression is widely applied to efficiently record image signals on the recording medium. There is.

As one of such image data compression methods, one that utilizes orthogonal transformation of image data is well known. This method divides gamma digital two-dimensional image data into blocks each having an appropriate number of samples, orthogonally transforms the numerical sequence consisting of sample values for each block, and through this transformation, energy is concentrated in a specific component. Therefore, components with high energy are encoded (ffi encoded) using a long code length υ1, while components with low energy 4゛- are coarsely encoded with a short code length, thereby reducing the number of codes per block. This is to avoid reducing the The above-mentioned orthogonal transformation includes Fourier (F).

Qosine'l
! transformation, Hadamard transformation, Karhunen-Lebe
−1, o @ v e ) conversion, bar (Ll a a
r) transformation etc. are often used, and the above method will be explained in more detail using the Hadamard transformation as an example. First, as shown in FIG. 2, digital two-dimensional image data is divided into two blocks in a predetermined one-dimensional direction to form blocks. When the two sample values x (0) and 1)=X(0). Therefore, this orthogonal coordinate system is transformed by 45° as shown in FIG. 3 to define a new y(◯)-1(1) coordinate system. In this coordinate system, y(0) indicates the low frequency component of the original image data of transformation na, and y(0
) is a slightly larger value than x (0) and x (1) (approximately f
2 times), but it is distributed only in a very narrow range near the yxx and Ly(0) axes, which indicate the high frequency components of the −5 original image data. So, for example, the above ×(0), x
If we say that encoding (1) requires a code length of 7 bits each, y(0) requires a 7-bit or 8-bit code length, while y(1) requires, for example, about 4 bits. Encoding can be performed with one code length, and the code length per block is eventually reduced, realizing image data compression.

As described above, one block is composed of two image data f.
The following orthogonal transformation has been explained, and as the order increases, the tendency for energy to be concentrated in a specific component becomes stronger, and the effect of reducing the number of pits can be increased. In general, the above transformation can be performed by using an orthogonal function matrix, and in the extreme, if the same right function of the target image is chosen as the orthogonal function matrix, the transformed image becomes its eigenvalue matrix, and the pair of matrices This means that the original image can be expressed using only the angular components. Also, in the above example, the image data is grouped into blocks only in one dimension, but this block may also be composed of several pieces of image data arranged in two dimensions, and in the case of In the case of B, a more remarkable bit loss reduction effect can be obtained than in the case of B.

The transformed data subjected to 1r7 in the above-mentioned two-dimensional orthogonal transformation are arranged in the order of the orthogonal functions used for transformation within each block (the number that crosses 0). Since this sequence corresponds to the spatial frequency, each converted data is arranged in frequency order in the vertical and horizontal directions, as shown in FIG. Therefore, a relatively long code length is assigned to the converted data that carries the low frequency component (the data on the upper left side of Figure 4) (the above-mentioned one-dimensional
(corresponds to assigning a long code length to y (0) in the second-order orthogonal transformation), transform data (fourth
The code length per block is reduced by assigning a relatively short code length v1 to the data on the lower right side of the figure or by truncating the code length v1.

The above-mentioned truncation of the converted data is performed by so-called zone sampling, as is well known.

In other words, as shown in FIG. 4, the converted data are arranged vertically and horizontally in order of frequency within each block B, so the arc R centered on the converted data y(1, 1) of sequence O
The converted data y of the same frequency component are lined up in ten rows.

If, for example, the converted data y in the shaded zone Z in FIG. 4 is sampled and only the sampled converted data y is converted to g, significant data compression can be achieved.

However, since the transformation data distribution state within the blocks B and B varies from image to image and from block to block, if the above-mentioned zone Z is fixed, the conversion that is useful for image reproduction in a certain block is Problems arise in that the data is truncated and the quality of the reproduced image deteriorates, or on the other hand, in a certain block, less useful converted data is also sampled, reducing the image data compression effect. In order to solve this kind of lz problem, conventionally, multiple sampling zone patterns are used, an appropriate zone pattern is selected based on the converted data distribution state of each block, and the selected pattern is applied. It is also being considered to perform zone sampling. However, in such a method, there is a practical limit to the number of zone patterns that can be prepared, and therefore, it is natural that zone sampling may be performed based on a zone pattern that is not very appropriate.

(Object of the Invention) The present invention has been made in view of the above circumstances, and it is possible to maximize the effect of compressing image data on any block of image data without impairing the quality of the raw image. It is an object of the present invention to provide an orthogonal transform encoding method for image data that can improve the image data.

(Constitution of the Invention) The orthogonal transform encoding method of image data of the present invention sequentially converts the characteristic values corresponding to the absolute values of the transform f-ta in each block from the high frequency component side to the low frequency component side. In each block, only the transformed data on the lower frequency side than the transformed data when K+ reaches a predetermined value is subjected to zone limp ringing and encoded. It is something.

The second orthogonal transform encoding method of image data of the present invention is basically the same as the first method described above, in which the transform data is lower in frequency than the transform data when the above-mentioned addition value reaches a predetermined value. Although only the data is zone sampled, the absolute (1α) of the J-transformed data that shows the frequency component is truncated (not sampled at all) when the absolute value of the J-transformed data is small as a whole (it is not sampled at all), but the area is unnecessarily low frequency. Specifically, a limit frequency is set in advance, and if the frequency when the above added value reaches a predetermined value is lower than this limit frequency, this limit frequency is eIJ, Sushi All the converted data on the low frequency side is sampled.

, (The third method of orthogonal transformation of image data of the present invention, etc. is basically the same as the first method described above, but is based on the converted data when the above-mentioned correction value reaches a predetermined value. Although it samples only the converted data on the low frequency side, it prevents the occurrence of density steps at the boundaries of the gray scales (so-called black distortion) in the reproduced image.
Specifically, when the difference between the sampling areas in adjacent blocks is less than a predetermined value, the numbering areas are corrected so that the difference is within a predetermined value. It is something.

As the above-mentioned Iζ "characteristic value corresponding to the absolute value of the converted data", the absolute value of the converted data, the square value of the converted data, etc. can be used.

(Function) When converted data that is not very important for image reproduction is relatively widely distributed in the high frequency region on the lower right side in FIG. Conversion data where the additive value of the characteristic value reaches a predetermined value exists on the lower frequency side. Therefore, in this case, according to the above method, the zone sampling area is set to be relatively narrow, and much of the less important conversion data is discarded, thereby increasing the image data compression effect.

On the other hand, when there are many converted data with large absolute values in the high frequency region shown in Figure 4 (they are important to some extent for image reproduction), the converted data where the sum of the above characteristic values reaches a predetermined value is , exist on the higher frequency side.

Therefore, in this case, the zone sampling area is set relatively wide, and although the image γ-data compression effect is somewhat reduced, the converted data for image reproduction is not discarded.

(Embodiment IIIi) Hereinafter, the present invention will be described in detail based on the embodiments shown in the drawings.

FIG. 1 schematically shows an equipment for implementing the first orthogonal transform encoding method of image data of the present invention. Image data (original image data) X representing a halftone image is first passed through a preprocessing circuit 10 and undergoes preprocessing such as smoothing to remove noise and other steps to increase data compression efficiency. The image data X that has undergone this preprocessing is passed through the orthogonal transformation circuit 11 and first undergoes two-dimensional orthogonal transformation. This two-dimensional orthogonal transformation is performed, for example, as shown in FIG. 5, as shown in FIG. ! ! This is performed on a rectangular block 81H of (number of pixels) MXN. For example, the aforementioned Hadamard transform is used as the orthogonal transform for Itaoka. Since the Hadamard transform has a transform matrix consisting of only 1-1 and -1, it can be executed by a simpler transform circuit compared to other orthogonal transforms. Furthermore, as is well known, two-dimensional orthogonal transformation can be reduced to one-dimensional orthogonal transformation. In other words, a one-dimensional orthogonal transformation is applied to the image data regarding the MXNij elements in the two-dimensional block I3 in the vertical direction, and then a 1-dimensional orthogonal transformation is applied to the obtained transformed data of MXN in the horizontal direction.
Two-dimensional orthogonal transformation is performed by applying dimensional orthogonal transformation. Note that the order of vertical and horizontal conversion may be 1 or reversed.

Transformed data y transformed by 1q by the above two-dimensional orthogonal transformation
As shown in Figure 4, in each block B, the functions that are the basis of the L orthogonal transformation (for example, the Walshlfl number for Hadamard's 9 transform, the trigonometric function for the Fourier transform, etc.) are arranged in the order of sheerancy. Arranged vertically and horizontally. As mentioned above, this sequence corresponds to the spatial frequency, so the converted data y are arranged vertically and horizontally in order of spatial frequency (in order of density of the Disol component of the image) in the above block B. to t. In addition, in this Figure 4,
Conversion data y in the top row and leftmost column (1, 1> is the sequence
This conversion data y(1,1) corresponds to O (ze[ko), and as is well known, this converted data y(1,1) indicates the average image sharpness within block B.

The converted data y arranged in this manner is sent to the encoding circuit 12 and encoded as shown in FIG.

This encoding circuit 12 encodes each 9-conversion data y in block B with a code (length and number of bits) according to an allocation bit allocation table created as described later.

For example, as shown in Fig. 6, the bit allocation table No. 1 allocates a number to a unique pin 1 for each sequence, and as mentioned above, the converted data y has energy concentrated in low frequency components. Therefore, by giving a relatively long code length to the low frequency component with high energy, and giving a relatively short code length to the low energy frequency component,
The number of bits required per block B is reduced and image data compression is achieved. Furthermore, as shown in Fig. 6, by allocating 0 bits to the converted data y that carries a particularly high frequency component and truncating it, and sampling only the other converted data y (so-called zone limp ring), the image data The compression effect is further enhanced.

However, if the above-mentioned zone limp ring area is too wide, it will be impossible to sufficiently increase the image data compression effect, and on the other hand, if the zone sampling area is too narrow,
Much of the conversion data y useful for image reproduction will be discarded. Hereinafter, features of the method of the present invention for appropriately setting the zone sampling area in accordance with the distribution state of the transformed data y, which differs from image to image and from block to block, will be explained.

The bit allocation determining unit 20 connected to the encoding circuit j'812 as shown in FIG. n1J is applied sequentially toward the low frequency component side. In other words, this addition is as shown by arrow 1-1 in Figure 4.
Sequence: Starting from the largest converted data y (M, N), the absolute value of each converted data y is added along a path along an arc R centered on the converted data y (1° 1). Then, when this additional value reaches a predetermined value at a certain converted data y, the bit allocation determining unit 20 allocates a code length of 1 bit or more to only the converted data y on the lower frequency side than that converted data y. , O(+:!mouth)pi]- is assigned to other conversion data y. In other words, if we explain using FIG. 4, the conversion data when the above predetermined value is reached is yO
-4, the arc R passes through this conversion data yO
0 bit is assigned to the converted data y1 and J along the arc RO concentric with , and to the converted data y located in the lower right area than the converted data y1 and J.

Jj The number of bits to be allocated to each converted data y given a code length of 1 bit or more may be determined according to a pre-stored allocation table, or it may be determined based on each actual converted data. Check the value of y and decide accordingly. Bit allocation determining unit 2 described above
Process 1 for determining bit allocation in 0! I! The flow is shown in Figure 7 for easy understanding.

The information Q indicating the bit allocation determined as described above is sent to the encoding circuit 12, and the encoding circuit 12 encodes each converted data y according to this bit allocation information. At this time, as described above, the /C converted data y to which 0 bit is assigned is truncated, and only the converted data y given a code length of 1 bit or more is sampled and encoded.

Here, the area of the transformed data y to which the O bit is assigned was determined as described above, so the image is reproduced in the high frequency fJ region within block B. If the target area is widely distributed, the zone sampling area is set to be relatively narrow. Therefore, the less useful transformation data y
, and the required number of bits per block B is greatly reduced. On the other hand, in block B, if the conversion data y (which is relatively unimportant for image reproduction, that is, the absolute value is large to some extent) is widely distributed up to the high frequency region, the zone sampling region is set relatively wide. Therefore, in this case, the required amount per block B (7 bit losses is higher than in the above case, but
Conversion data y useful for image reproduction will not be truncated.

The image data 9(y) encoded as described above and the position information of the arc RO are recorded on a recording medium (image file) such as an optical disk or a magnetic disk by the recording/reproducing device 13 as shown in FIG. be done. As mentioned above, this image data a (y) has been significantly compressed with respect to the original image data X, so a large amount of images are recorded on a recording medium such as an optical disk. During image reproduction, this image data g (y) and the positional information of the arc RO are read out from the recording medium, and the decoding circuit 1
d3 and is decoded into the converted data y. The converted data y decoded in this way is sent to the inverse conversion circuit 15,
It undergoes an inverse transformation to the two-dimensional orthogonal transformation. As a result, the original image data X is restored, this original image data X is sent to the image reproducing device 16, and the image carried by the data X is reproduced.

Next, an embodiment of the second method of the present invention will be described.

This second method is performed using an apparatus having basically the same configuration as the apparatus shown in FIG. It can be done.

However, in this case, the bit allocation determining section 2( ) is provided with a storage means as described below. The following is the allocation decision for Pi1]! This bit allocation determination process will be explained with reference to FIG. 8 showing the flow of step 11 and FIG. 10.

A predetermined limit frequency F' as shown in FIG. This arc R' defines the minimum zone sampling area that does not reduce the quality of the reproduced image.

If the sampling area is narrowed down to the lower frequency range side than this arc R', there is a risk that the quality of the reproduced image will be impaired. (
7 Such a circular arc R' can be determined experimentally and empirically by J
3 <I can do it.

The bit allocation determining unit 20 converts the absolute value of each converted data y into converted data y(M
, N>, and when the added value reaches a predetermined value, the converted data y. seek.

Next, the bit allocation determining section 20 determines which of the frequency FO corresponding to the converted data yO and the limit frequency F' is higher. Conversion data V.

When the frequency "0" corresponding to "0" is higher than the limit frequency F' (as shown in FIG. 10), the bit allocation determining unit 2
0 is an arc n that passes through the converted data y0 with the converted data Y(1, 1) as the center, as in the first method described above.
The region on the lower frequency side than o is defined as a zone sampling region. On the contrary, the limit frequency aF is higher than the frequency FO.
' is higher, that is, in FIG. 10, J3 is the converted data y.

is closer to the converted data y (1, 1>) than the converted data y' on the arc R', the bit allocation determining unit 20 performs zone sampling on the region on the lower frequency side than the arc R'. The bit allocation to each converted data y in the thus determined zone sampling area 43 is performed in the same manner as in the first method.

The encoding circuit 12 encodes the converted data y according to the bit allocation determined as described above, but in this second method, since the zone sampling area is determined as described above, the high frequency area Even if the converted data y whose absolute value is small is widely distributed, the zone sampling area will not be unnecessarily narrowed. In other words, the zone sampling area is at least +iff
This is a region on the lower frequency side than the arc R'. By doing this, the converted data y useful for reproducing both meetings will not be truncated.

Next, an embodiment of the third method of the present invention will be described.

This third method can also be implemented in an apparatus having the same structure as the loading shown in FIG. 1 above.

However, in this case, the bit allocation determining section 20 is provided with a storage means for storing the zone sampling area of the adjacent block B. The bit allocation determining process will be explained below with reference to FIG. 9 and FIG. 11, following the flow of the bit allocation determining process. The bit allocation determination unit 20 determines the zone sampling area J3 and bit allocation in each block B in the same manner as in the first method, but the bit allocation determination unit 20 determines the zone sampling area J3 and bit allocation in each block B. A correction process is also performed to keep the difference within a predetermined value. That is, for example, as shown in FIG.
) The zone limp ring area of block B n-+ is on the lower frequency side than the arc R++-+, and the block B11 is The zone sampling area is an arc Rn
Suppose that it is a region on the lower frequency side than . When determining the bit allocation for block Bn, the bit allocation determining unit 20 stores information indicating the circular arc Rn-+ in the storage means 7 and stores the information indicating the circular arc Rn-+.
i, find the radius of this arc R and the radius of the arc fln.

Then, the pill-allocation determining unit 20 determines the
If the difference in fY exceeds a predetermined value fal'lr, then the arc RI+ is modified so that the difference becomes r.

FIG. 11 shows that the radius of arc Rn is r to the radius of arc Rn-1.
(Since it exceeds +11, it is shown that the radius is corrected to the arc Rn', but if the radius of the arc 1111 is less than 11a, which is the radius of the arc R+1-+ minus r. Even in the case where the bit allocation determination unit 20
is the arc Rn' modified in this way or the arc R'
1 (if no modification is required) defines the zone layer f sampling area and determines the bit allocation in that area. Then, the encoding circuit 12 converts the converted data y into H, p; according to this p1-allocation. Further, the information indicating the arc Rn or Rn' is stored in the storage means in place of the information indicating the arc R of the block 11n-+, and the zone sampling of the block 13 r++1 is then subjected to orthogonal transformation. Used to modify the area if necessary.

By making sure that the zone sampling areas of adjacent blocks do not deviate greatly as described above, it is possible to prevent so-called block distortion, where an m-degree step difference occurs at the boundary between each block in the reproduced image. It can be prevented.

In the embodiment described above, each block has 3
3, the absolute value of the conversion filter 2-ta is added and the added value is compared with a predetermined value.The characteristic value thus added is the sign corresponding to the above absolute value. Therefore, for example, 2 of the converted data y
The multiplier value or the like may be added. Also, as mentioned earlier, the lower the frequency side of the converted data y, the more important it is for image reproduction, so we use Kawamitsuke ( It would be more preferable if you use the value you have copied.In other words, if you do it like that,
When transformed data y with small absolute values are widely distributed in a high frequency region, it is possible to prevent the zone sampling region from being unnecessarily narrowed and the same transformed data y being discarded during image reproduction.

In addition, the predetermined value to be compared with the addition value of the special value may be determined in advance by an optimal value, or
The characteristic value of 1 is multiplied by 9B, and this Jllii product 11t1
It may be a value obtained by multiplying
, 1), all other characteristic values in the block 13 may be accumulated, and this cumulative value may be multiplied by a predetermined ratio.

(Effects of the Invention) As explained in detail above, the image edge-to-edge orthogonal transform encoding method of the present invention performs orthogonal transform of image data block by block (zone limpling of q transform data). The sampling area can be set to a particularly small size after collecting and numbering useful conversion data for reproduction.Therefore, according to the method of the present invention, the image 'tτ of the image reproduced through data compression cannot be degraded. This makes it possible to significantly increase the data compression rate.

In particular, in the second orthogonal transform encoding method of image data of the present invention, J3 sets a lower limit of the zone sampling area to prevent the area from becoming unnecessarily small, so that the encoding of the transformed data At this time, it is possible to prevent the image quality of the reproduced image from deteriorating due to the nm conversion data being discarded.

Furthermore, in the third method of orthogonal transformation rq of image data of the present invention, since the sampling areas of adjacent blocks are not far apart, block distortion can be prevented from occurring in the reproduced image. .

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a schematic depiction of an apparatus for carrying out the method of the present invention; FIGS. 2 and 3 are explanatory diagrams for explaining orthogonal transformation according to the present invention; FIG. 5 is an explanatory diagram explaining the block division of an image according to the present invention. FIG. 6 is an explanatory diagram explaining the bit allocation table according to the present invention. Figures 7.8 and 9 respectively illustrate the invention in Section 1.2 J.
10 and 11 are flowcharts showing the flow of the bit allocation determination process according to methods J and 3, respectively.
FIG. 6 is an explanatory diagram illustrating zone sampling in the third method. 11... Orthogonal transform circuit 12... Encoding circuit 2
0...Bit allocation determining unit F'...Limit frequency 11, Bn-+, Bn, Bnn=1 RN per image data pool Raz Rn-1z Rn %
Rn'...A circular arc r covering the center of the conversion data of sequence 〇...Difference in the radius of the arc Conversion data when the value is reached g (y) ... Encoded image data Figure 4 Figure 6 Figure 5 Figure 10 (Voluntary) Mr. Showa, Director General of the Special Office of the City of N. June 4, 1961 1, Indication of matter f1 Patent Application No. 1983-90535 2, Name of invention Method for converting image data into 4-3 orthogonal transformation code 3, Relationship with famous matter f1 to be corrected Appointment of patent applicant Address: 210 Nakanuma, Minamiashigara City, Kanagawa Prefecture Name:
Fuji Photo Film Co., Ltd. 4th generation Iq! 5-1 Roppongi, Minato-ku, Tokyo: l 2 ffl 1-6,
Number of inventions increased by amendment None 7. Target of amendment: Drawing 8. Contents of amendment: Hand-drawn drawings have been amended to inked drawings. (No change in content)

Claims (8)

[Claims] (1) After applying orthogonal transformation to the two-dimensional image data for each block based on a predetermined orthogonal function, the transformed data arranged in the sequence of the orthogonal function within the block after undergoing this transformation is uniquely In the orthogonal transform encoding method for image data that is encoded with a code length of In the block, only the converted data on the lower frequency side than the converted data when this added value reaches a predetermined value is sampled, and each of the sampled converted data is encoded with a unique code length. Orthogonal transform encoding method for image data. (2) The orthogonal transform encoding method for image data according to claim 1, wherein the characteristic value is an absolute value of the transformed data. (3) The orthogonal transform encoding method for image data according to claim 1, wherein the characteristic value is a square value of the transform data. (4) The image data according to any one of claims 1 to 3, wherein the characteristic value is weighted by the reciprocal of the spatial frequency carried by the corresponding transformation data. Orthogonal transform encoding method. (5) The predetermined value is a value obtained by multiplying a cumulative value of all characteristic values in the block by a predetermined ratio, according to any one of claims 1 to 4. An orthogonal transform encoding method for image data. (6) The predetermined value is sequence 0 within the block.
Claims 1 to 4, characterized in that the value is the cumulative value of all other characteristic values except for the characteristic value related to conversion data of (zero) multiplied by a predetermined ratio. An orthogonal transform encoding method for image data. (7) After applying orthogonal transformation to the two-dimensional image data for each block based on a predetermined orthogonal function, the transformed data arranged in the sequence of the orthogonal function within the block after undergoing this transformation is uniquely In the orthogonal transform encoding method for image data that is encoded with a code length of In the block, only the frequency corresponding to the converted data when this added value reaches a predetermined value and the converted data on the lower frequency side than the higher frequency of the pre-specified limit frequency are sampled. An orthogonal transform encoding method for image data, characterized in that each transformed data is encoded with a unique code length. (8) After applying orthogonal transformation to the two-dimensional image data for each block based on a predetermined orthogonal function, the transformed data arranged in the sequence of the orthogonal function within the block after undergoing this transformation is uniquely In the orthogonal transform encoding method for image data that is encoded with a code length of In the block, only the converted data on the lower frequency side than the converted data when this added value reaches a predetermined value is sampled, and each of the sampled converted data is encoded with a unique code length. An orthogonal transform encoding method for image data, characterized in that when the difference between the sampling areas in the blocks exceeds a predetermined value, the sampling areas are modified so that the difference falls within a predetermined value.