JPH114438A - Image compressor and camera - Google Patents

Abstract

PROBLEM TO BE SOLVED: To compress an image by predicting a quantization error, so as to reduce an image deterioration due to quantization and inverse quantization. SOLUTION: This image compressor 10 is provided with a quantization table generating circuit 20. The quantization table generating circuit 20 is provided with a statistical quantity storage section 21, a quantization error predict section 22, and a quantization table calculation section 24. A DCT processing circuit 12 applies discrete cosine transform(DCT) to image data, that is, a luminance signal and color difference signals Cb, Cr to convert the data into a DCT transform coefficient. The statistic quantity storage section 21 calculates sampling number data which is a statistical quantity with a close relation to a quantization error from the DCT transform coefficient for each spatial frequency, stores sampling number data of a prescribed number, and updates the data for each image input. The quantization error predict section 22 uses the sampling number data to conduct quantization error prediction. The quantization table calculation section 24 calculates quantization tables Qya, Qca, based on the quantization error predict result by the quantization error predict section 22, and outputs them to a quantization-processing circuit 14.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image compression apparatus and a camera for compressing information of a color still image according to the JPEG algorithm.

[0002]

2. Description of the Related Art JPEG (Joint Photographic Expert Group) has recommended a standardized algorithm for encoding (compressing) a high-resolution image and transmitting and receiving information via a communication transmission line. In the JPEG algorithm process, original image data is decomposed into components on a spatial frequency axis by a two-dimensional discrete cosine transform (hereinafter, referred to as a two-dimensional DCT), and each data represented on the spatial frequency axis is quantized. Significant information compression is performed by encoding each quantized data. The compressed image data is expanded by decoding and inverse quantization.

[0003] JPEG recommends a predetermined quantization table for quantization or inverse quantization. This D
Image compression by CT and quantization is an efficient image compression for a wide range of subjects, but is an irreversible method involving an error between the original image data and the expanded reproduced image data.

[0004]

In the quantization or inverse quantization of image compression, regardless of the variety of subjects,
Since a single quantization table is used, there is a problem in that reproduced image data greatly deteriorates depending on a subject.

SUMMARY OF THE INVENTION In view of the above problems, an object of the present invention is to provide an image compression apparatus and a camera which predict a quantization error and facilitates image compression with less image deterioration due to quantization and inverse quantization. It is.

[0006]

An image compression apparatus according to the present invention quantizes an orthogonal transform coefficient corresponding to original image data obtained from a photographing optical system using a quantization table, and quantizes the quantized orthogonal transform coefficient. It is characterized by comprising a quantization means to be obtained and a quantization table creation means for updating the quantization table based on the orthogonal transform coefficient.

In the image compression apparatus, preferably, the quantization table creating means has an absolute value of an error between the inversely quantized orthogonal transform coefficient and the orthogonal transform coefficient obtained by inversely quantizing the quantized orthogonal transform coefficient, which is larger than two. A storage unit is provided which counts the number of samples of a large orthogonal transform coefficient for each orthogonal transform coefficient, and stores data of the number of samples (sample number data) in each image for a predetermined number of images.

In the image compression apparatus, preferably, the sample number data of the oldest image stored in the storage unit is deleted every time the sample number data of a new image is input to the storage unit. The number of samples data is updated.

In the image compression apparatus, preferably, the quantization table creating means includes a quantization error prediction section for predicting an error based on a predetermined number of sample data, and a prediction error obtained from the quantization error prediction section. A quantization table calculation unit that calculates a quantization table based on the value and outputs the quantization table to a quantization unit.

In the image compression apparatus, preferably, the quantization table creating means creates a quantization table based on a predetermined number of sample number data including the newly input sample number data. Preferably, the quantization table creating means creates the quantization table based on a predetermined number of sample number data input before the newly input sample number data.

Further, the camera according to the present invention quantizes orthogonal transform coefficients corresponding to original image data obtained from the photographing optical system by using a quantization table to obtain a quantized orthogonal transform coefficient. A quantization table creating unit that updates the quantization table based on the transform coefficient,
The quantization table creation means corresponds to the image data of one image, and the absolute value of the error between the inversely quantized orthogonal transform coefficient and the orthogonal transform coefficient obtained by inversely quantizing the quantized orthogonal transform coefficient is greater than 2. A sample number data storage unit that counts the number of samples of the orthogonal transform coefficient for each orthogonal transform coefficient and stores data of the number of samples in each image (sample number data) for a predetermined number of images; A quantization error prediction unit that predicts an error based on a predetermined number of sample number data stored in the quantization error prediction unit, and calculates a quantization table based on a prediction error value obtained from the quantization error prediction unit. Recording means for recording on a recording medium compressed image data compressed using the quantization table created by the quantization table creating means , And a quantization table storage means for storing the quantization tables created.

[0012] In the camera, preferably, the recording means records the compressed image data in a first area of the recording medium, and a quantization table used for quantization is recorded in a second area of the recording medium.

[0013] In the camera, preferably, the quantization table creating means uses the quantization table stored in the table storage means for quantization, and after the recording of the compressed image data on the recording medium is completed, A quantization table is created based on the compressed image data and the corresponding sample number data, and the quantization table is updated by storing the created quantization table in the quantization table storage means.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of an image compression apparatus and a camera according to the present invention will be described below with reference to the accompanying drawings.

FIG. 1 shows a schematic configuration of a camera according to an embodiment of the present invention. A subject image (still image) of the subject S obtained through the photographing optical system 11 is separated into red (R), green (G), and blue (B) images by the color separation optical system 13, and is composed of, for example, a CCD. An image is formed on three image sensors 15. That is, R, G, and B images are formed on the respective image sensors 15. These R,
The G and B image signals are subjected to predetermined processing in a signal processing unit 17, converted into luminance data Y and color difference data Cb and Cr, and input to an image memory 19. Image memory 19
Are divided into mutually independent memory areas for storing the luminance data Y and the color difference data Cb and Cr, respectively, and each memory area has a storage capacity for one image. The luminance data Y and the color difference data Cb and Cr are input data (original image data) to the image compression device 10.

The luminance data Y or the color difference data Cb,
Cr is subjected to two-dimensional DCT in the DCT processing circuit 12, and is converted into DCT coefficients for each spatial frequency. Based on these DCT coefficients, the quantization table creation circuit 20
Thus, a quantization table Qya for luminance data and a quantization table Qca for color difference data are created, and are output to the quantization processing circuit 14. The quantization table creation circuit 20 is provided with a statistic storage unit 21, a quantization error prediction unit 22, and a quantization table calculation unit 24, and the configuration and operation thereof will be described later in detail.

The luminance data Y or the color difference data Cb,
The DCT coefficient of Cr is quantized by the quantization processing circuit 14 using the quantization tables Qya and Qca created by the quantization table creation circuit 20, and the quantized DCT
Converted to coefficients. The quantized DCT coefficients of the luminance data Y or the color difference data Cb and Cr are encoded (compressed) by an encoding processing circuit (not shown) according to the JPEG algorithm, and are recorded as compressed image data in the compressed image data recording area M1 of the recording medium M. Is done. The quantization tables Qya and Qca used in the quantization processing circuit 14
Is recorded in the table recording area M2 of the recording medium M.
The recording medium M is, for example, a removable IC card.

FIG. 2 shows a conventional quantization table recommended by JPEG. FIG. 2A shows a quantization table Qy for the luminance value Y, and FIG. 2B shows a quantization table Qc for the color differences Cb and Cr.

Referring to FIG. 3, quantization and inverse quantization will be described. FIG. 3 shows, as an example, original image data of 8 × 8 pixels, that is, a DCT coefficient F _vu of luminance data Y,
A quantized DCT coefficient R _vu and an inversely quantized DCT coefficient F ′ _vu ,
And a quantization table Qy. The subscripts v and u are
The vertical and horizontal positions when the 64 DCT coefficients are displayed in the form of an 8 × 8 matrix are shown, and the subscript v is 0, 1, 2,. . . 7 and the subscript u is 0, 1, 2,. . . 7

The original image data is converted into 8 × 8 = 64 DCT coefficients F _vu by two-dimensional DCT in the DCT processing circuit 12. Since two-dimensional DCT is known,
It will not be described in detail here.

Of the 64 DCT coefficients, the position (0,
0) is a DC (direct current) component,
The remaining 63 DCT coefficients F _vu are AC (alternating current) components. The AC component is calculated from the coefficient F01 or the coefficient F10 to the coefficient F
Towards 77, it shows how much higher spatial frequency components are in the original image data of the 8 × 8 pixel block. The DC component represents the average value (DC component) of the pixel values of the entire block of 8 × 8 pixels. That is, each DCT coefficient F _vu corresponds to a predetermined spatial frequency.

Using a quantization table Q _vu , a DCT coefficient F
_The equation for quantizing _vu is defined by equation (1). Round in this equation means approximation to the nearest integer.
That is, the division of each element of the DCT coefficient F _vu and the quantization table Q _vu and the rounding-off result in the quantization DC
The T coefficient R _vu is determined.

R _vu = round (F _vu / Q _vu ) (1) {0 ≦ u, v ≦ 7}

As described above, the quantized DCT coefficient R _vu obtained by the quantization processing circuit 14 is encoded for each DC component and AC component using, for example, Huffman coding conforming to JPEG, and compressed image data Is recorded on the recording medium M. Since the Huffman coding is conventionally known, a detailed description is omitted.

In order to decompress the coded compressed image signal and display it on the screen, decoding, inverse quantization, two-dimensional DC
Processing of the inverse transformation of T (hereinafter referred to as two-dimensional IDCT) is required. This decoding is the reverse of the operation of Huffman coding, and is well known in the art and will not be described in detail. The quantized DCT coefficients obtained by the decoding are inversely quantized using the quantization tables Qy and Qc used for quantization, and are converted into inversely quantized DCT coefficients F ′ _vu . These inversely quantized DCT coefficients F ′ _vu are inversely transformed by two-dimensional DCT (hereinafter, 2
(Referred to as dimension IDCT), and converted into luminance data Y ′ and color difference data Cb ′ and Cr ′, respectively. Since the two-dimensional IDCT is also known, it will not be described in detail here.

For example, the DCT coefficient F00 (= 2
61) is obtained by Expression (1) using the quantization coefficient q00 (= 16), the quantization DCT coefficient R00 becomes 16, and
Q00 (= 16) is added to the quantized DCT coefficient R00 (= 16).
, The inverse quantized DCT coefficient F'00 (= 16 ×
16 = 256). The DCT coefficient F00 (= 26
The difference value i00 (256-261 = -5) between 1) and the inversely quantized DCT coefficient F'00 (= 256) is defined as a coefficient value error. In this way, in the quantization, the remainder of the division is rounded, so that a coefficient value error i occurs in the expansion (inverse quantization) of the compressed image data. This is the cause of the quantization error.

[0027] FIG. 4 is a diagram illustrating the DCT coefficients F _vu, a coefficient value error i _vu for each coefficient of the inverse quantized DCT coefficients F _'vu after quantization and inverse quantization. The present embodiment predicts the coefficient value error i _vu caused by the quantization or inverse quantization in this way, calculates each quantization coefficient q _vu , and creates the quantization tables Qy and Qc to perform the quantization. A configuration is provided to reduce the error.

Referring to FIGS. 1 and 5, the quantization table creation processing in the quantization table creation circuit 20 will be described. For example, when one piece of image data of 8 × 8 pixels having M blocks is input (hereinafter referred to as the latest image data), first, in a signal processing circuit (not shown), luminance data Y and chrominance data Cb, Cr Is converted to
The DCT is performed in the DCT processing circuit 12, and the DCT is performed.
It is converted into a coefficient and output to the statistic storage unit 21.

In the statistic storage unit 21, the statistic of the prediction error value is calculated from the v × u × M DCT coefficients with respect to the DCT coefficient of the latest image data. Statistics storage 2
1 is a predetermined number (N) obtained immediately before the latest image data.
The number of samples at each frequency in the image data (hereinafter referred to as sample number data) is recorded. When new sample number data is input, the oldest sample number data is deleted and the latest sample number data is deleted. The data is rewritten to new sample number data of N sheets including numerical data.

The quantization error prediction section 22 predicts a quantization error from the sample number data of N pieces obtained from the statistic storage section 21. The quantization table calculator 24 calculates the quantization tables Qya and Qca based on the result of the quantization error prediction by the quantization error predictor 22, and outputs it to the quantization processing circuit 14.

The quantization error prediction in the quantization error prediction section 22 will be described in detail. If the quantization coefficient in the quantization table Qy is q _vu , the DCT coefficient F _vu and the inverse quantization D
The range of the coefficient value error i _{vu from} the CT coefficient F ′ _vu is represented by the following equation (2).

−q _vu / 2 ≦ _ivu ≦ q _vu / 2 (2)

For example, if the quantization coefficient q _vu = 6, the coefficient value error i _vu corresponds to any of -3, -2, -1, 0, 1, 2, and 3. The range of the coefficient value error i _vu is the quantization coefficient q
Expands as _vu grows.

Next, the coefficient value error i _vu of the M × N block is
It is predicted that the distribution is averaged within the range shown in the equation (2),
The predicted value rms _vu of the coefficient value error i _{vu is obtained} by the equation (3). In the equation (3), i is a coefficient value error i _vu , q
Indicates a quantization coefficient q _vu , and rms indicates a square root of a root mean square value of a coefficient value error i _vu at each spatial frequency, that is, an rms value. For example, if q = 6, rms = about 1.7795.

[0035]

(Equation 1)

FIG. 6 shows a correspondence table between the quantization coefficient q and the rms value. In the present embodiment, the luminance data Y is represented by 8 bits (256 steps), and the quantization coefficient q is 1 to 255.
To change. As is well shown in this table, as the quantization coefficient q increases, the range of the coefficient value error i also increases, so that the rms value increases.

In the present embodiment, the DCT coefficient F which satisfies a predetermined condition in all M × N blocks, for example, the absolute value of the DCT coefficient F _vu at each spatial frequency is larger than 2
The number of _vu is set to the number of samples E _vu (0 ≦ E _vu ≦ M × N). In this embodiment, when the absolute value of the DCT coefficient F _vu is greater than 2, the coefficient value error i _vu increases as the value of the quantization coefficient q _vu increases, and as the number of samples E _vu increases,
_Paying attention to the fact that the quantization error of the entire image increases, the prediction error total value G _vu is expressed by the following equation (4).

G _vu = rms _vu × E _vu (4)

If the prediction error total value G _vu is uniform for each spatial frequency, the quantization error of the entire image is reduced, and reproduced image data with high resilience at the time of image expansion is obtained. Therefore, in the present embodiment, first, using the lower spatial frequency, in particular, the property that image information concentrates on the DC component,
The total prediction error G00 of the DCT coefficient F00, which is the C component, is
It is obtained from the predetermined quantization coefficient q00 determined in advance by the equations (3) and (4). Other DCT coefficients F
_vu (v, u ≠ 0), that is, the prediction error total value G _vu of the 63 AC components is the same as the prediction error total value G ₀₀ of the DC component, and the corresponding prediction error rms ′ _vu is calculated by the equation (5). It is calculated and output to the quantization table calculation unit 24.

Rms' _vu = G 00 / E _vu (5) where u, v ≠ 0

Next, the quantization table creation processing in the quantization table calculation unit 24 will be described. The prediction error rms' _vu obtained by the above-described quantization error prediction is (6)
The prediction quantization coefficient q ′ _vu (v, u ≠ 0), which is an integer closest to q ′ that satisfies the expression (6), is substituted into the expression, and is obtained.

[0042]

(Equation 2)

For example, when processing 10 images of 480 × 720 pixels, the number M of blocks per image is 5,400.
And the total number of blocks M × N is 5400 × 10 = 54000.
At this time, the quantization coefficient q00 corresponding to the DC component is set to 6, and the sample number E00 in which the absolute value of the DCT coefficient F00 is larger than 2
Is 53940, the total prediction error value G00 = 1.7995 × 53940 = 95986.23 is obtained from the equations (3) and (4). Assuming that the AC component, for example, the sample number E77 in which the absolute value of the DCT coefficient F77 is larger than 2, is 31520, the prediction quantization coefficient rms'77 = 95986.23 ÷ 31520 = approximately 3.0 by equation (5).
452 is obtained, and a predicted quantization coefficient q′77 = 11 that satisfies the expression (6) is obtained.

Each predicted quantization coefficient q ′ _vu obtained as described above is scaled according to the compression ratio, and output to the quantization processing circuit 14 in the form of an 8 × 8 matrix Qya. In image compression, high compression can be achieved by increasing the quantization coefficient q, but image quality deteriorates. Conversely, reducing the quantization coefficient q improves image quality, but increases the amount of data after compression. The image quality is controlled by multiplying the quantization coefficient q by a predetermined compression ratio.

FIG. 7 shows an example of the created quantization table. FIG. 7A shows a quantization table Qya for the luminance value Y.
FIG. 7B shows a quantization table Qca for the color differences Cb and Cr. Compared to FIG. 2, there is a difference particularly in the corresponding quantization coefficient in the AC component.

Referring to the flowcharts of FIGS. 8 and 9,
The statistical processing in the statistic storage unit 21 will be described. 8 and 9, BLOCKS is a variable indicating the number M of blocks, and BA is equal to the total number of blocks to be processed (M × N). F _vu indicates a DCT coefficient, and the subscripts v and u change from 0 to 7, respectively. E _vu is a variable indicating the number of samples.

First, reference is made to FIG. In step S102, the initial values of the variables BLOCKS and E _vu are set to 0, respectively. In step S104, the initial value of the variable v is set to 0, and in step S106, the initial value of the variable u is set to 0. In step S108, it is determined whether the absolute value of the DCT coefficient F _vu is larger than 2. DCT coefficient F
_If the absolute value of _vu is larger than 2, the number of samples E _vu is incremented by one in step S110, and the value in step S1 is increased.
Proceed to 12. If the absolute value of the DCT coefficient F _vu is not larger than 2, the process proceeds to step S112, where the variable u is incremented by one.

In step S114, it is determined whether or not the variable u is 8, and if u is not 8, ie, 7 or less,
The processing is executed again from step S108. If u = 8, the process proceeds to step S116, where the variable v is incremented by one. Similarly, in step S118, it is determined whether or not the variable v is 8, and if v is not 8, ie, 7 or less,
The processing is executed again from step S106. If v = 8, the process proceeds to step S120.

In step S120, the variable BLOCKS
Is incremented by one, and it is determined in step S122 whether the variable BLOCKS is the total block number BA. If BLOCKS is not BA, step S1
04, and if BLOCKS = BA, the process proceeds to step S124.

As described above, in the matrix F _vu of the DCT coefficients of 8 × 8, the sample number E _{vu of} the top horizontal row is counted sequentially from the left, and the sample number E _vu of the next horizontal row is sequentially counted.
Is counted. That is, in the process of FIG. 8, the number of samples E _vu whose absolute value of the DCT coefficient F _vu is larger than 2 is counted in each of the 64 frequency components.

Next, reference is made to FIG. In step S124, the initial value of the variable v is 0, in step S126, the initial value of the variable u is 0, and in step S128, the variable i representing the number of images
Is set to 0. In step S130, E
_vu [i + 1] is converted to E _vu [i]. For example, if E [1] is E
[0] and E [6] becomes E [5]. Step S134
In, it is determined whether or not the variable i is equal to the number (N-1) one less than the number of images. If the variable i is not equal to (N-1), step S130 is executed again, and the variable i becomes (N-
If it is equal to 1), the process proceeds to step S136. Step S
At 136, E _vu [N-1] is set to E _vu , and step S
Proceed to 138.

In step S138, the variable u is incremented by one, and in step S140, it is determined whether or not the variable u is 8. If u is not 8, that is, 7 or less, the process is repeated from step S128. If u = 8, the process proceeds to step S142, and the variable v is incremented by one. Similarly, in step S144, it is determined whether or not the variable v is 8, and if v is not 8, that is, 7 or less, the process is repeated from step S126. If v = 8, the process ends.

For example, N = 10 sample number data (E
[0], E [1],. . . E [9]; the larger the numerical value in parentheses, the more recent the image data is.) Is stored in the statistic storage unit 21. When the latest sample number E is input to the statistic storage unit 21, the subscript i (1 ≦ i ≦ N-1) of the stored sample number data E [i] is lowered by one (step S130). , The sample size data is E
[0], E [1],. . . E [8]. Then, the latest sample number data E is updated to E [9] (step S13).
6). The updated E [0] to E [8] and E [9] are stored again in the statistic storage unit 21 and output to the quantization error prediction unit 22. Thus, in the processing of FIG.
The N sample number data E stored in the statistic storage unit 21 is constantly updated with the N sample number data E of the latest captured image.

The quantization error prediction processing in the quantization error prediction section 22 will be described with reference to the flowcharts of FIGS. 10 and 11, E _vu is a variable indicating the number of samples.

First, reference is made to FIG. Step S202
Then, the initial value of the variable E _vu is set to 0. Step S
At 204, the initial value of the variable v is set to 0, and at step S206, the initial value of the variable u is set to 0. In step S208, the initial value of the variable i is set to 0.

In step S210, the number of samples E _vu [i] at each spatial frequency is added, and step S21 is performed.
In step 2, the variable i is incremented by one,
Proceed to 214. In step S214, it is determined whether or not the variable i is N. If i is not N, step S210
Is re-executed from If it is determined that the variable i is N, the process proceeds to step S216. That is, through the processing from step S208 to step S214, N total sums of the number of samples E _vu at each spatial frequency are obtained.

In step S216, the variable u is incremented by one. In step S218, it is determined whether or not the variable u is 8, and if u is not 8, ie, 7 or less, the process is re-executed from step S208. If u = 8, the process proceeds to step S220, and the variable v is incremented by one. Similarly, in step S222, it is determined whether or not the variable v is 8, and if v is not 8, ie, 7 or less, the process is re-executed from step S206. If v = 8, the process proceeds to step S302.

As described above, in the processing of FIG. 10, the number of samples E of each image i counted in the flowchart of FIG.
The sum Evu of _vu [i] is obtained for each spatial frequency.

Next, reference is made to FIG. Step S302
Then, the initial value of the variable v is 0, and in step S304, the variable u
Is set to 0. In step S306, it is determined whether both the variables v and u are 0. Variables v, u
If both are 0, the process proceeds to step S308, where the variables v and u
Are not 0, the process proceeds to step S310.

In step S308, the rms value of q00, that is, rms00, is obtained using equation (3) based on the predetermined quantization coefficient q00. By multiplying this rms00 by the number of samples E00, the total prediction error G00 is obtained, and the process proceeds to step S312.

In step S310, the prediction error total value G
By substituting 00 and the number of samples E _vu into the equation (5), the rms value rms ′ _vu is obtained. Then, rms' _vu (6)
The predicted quantization coefficient q ′ _vu is obtained by substituting into the equation, and the process proceeds to step S312.

In step S312, the variable u is incremented by one, and in step S314, it is determined whether or not the variable u is 8. If u is not 8, that is, 7 or less, the process is re-executed from step S306. If u = 8, the process proceeds to step S316, where the variable v is incremented by one. Similarly, in step S318, it is determined whether or not the variable v is 8, and if v is not 8, ie, 7 or less, the process is re-executed from step S304. If v = 8, the process ends.

As described above, first, D is calculated by the quantization coefficient q00.
The total prediction error G00 of the C component is obtained, and the rms of the DC component is obtained.
00 is determined (step S308). Then, each rms value rms ' _vu of the AC component is obtained from each total sample number E _vu and the prediction error total value G00, and the prediction quantization coefficient q' _vu (v,
u ≠ 0) is calculated (step S310). That is, FIG.
In the processing of A, based on the total prediction error G00 of the DC component, A
63 predicted quantization coefficients q ′ _vu corresponding to each frequency component of the C component are determined.

In the first embodiment, a quantization table creation circuit 20 is provided.
A statistic storage unit that calculates, for each spatial frequency, sample number data, which is a statistic closely related to the quantization error, from DCT coefficients of the original image data, stores a predetermined number of sample number data, and updates each time an image is input 21. The image processing apparatus further includes a quantization error prediction unit 22 that predicts a quantization error using the sample number data, and a quantization table calculation unit 24 that obtains a quantization coefficient based on the quantization error and outputs a quantization table. ing. Accordingly, a single quantization table Qya optimal for a predetermined number of latest images can be created. Therefore, reproduced image data with a small quantization error can be obtained with the same amount of compressed image data.

FIG. 12 is a view showing a second embodiment of the present invention. The same components as those of the first embodiment are denoted by the same reference numerals as those of the first embodiment, and description thereof is omitted here. In the second embodiment, the quantization tables created by the quantization table creation circuit 20 are Q′y (for luminance data), Q′c
(For color difference data).

The quantization table storage section 30 is a memory in which, for example, default quantization tables Qy and Qc are written. D obtained from the DCT processing circuit 12
The CT coefficient F _vu is input to the quantization processing circuit 14 and also to the quantization table creation circuit 20. In the quantization processing circuit 14, default quantization tables Qy,
The quantization is performed using Qc. The quantized image data is encoded and recorded in the compressed image data recording area M1 of the recording medium M as compressed image data. The default quantization tables Qy and Qc are recorded in the table recording area M2 of the recording medium M.

Next, the quantization table creation circuit 20 calculates the quantization table Q ′ based on the quantization error due to the newly input DCT coefficient and the past quantization error recorded in the statistic storage unit 21. y and Q′c are created and output to the quantization table storage unit 30. The quantization table storage unit 30 rewrites the quantization tables Q′y and Q′c obtained from the quantization table creation circuit as default quantization tables Qy and Qc.

Therefore, the quantization tables Qy and Qc used for the quantization are created based on the image data previously input. In the second embodiment, as in the first embodiment, the quantization errors are predicted from the DCT coefficients of the respective images, and the quantization tables Qy and Qc are created by updating the statistics of the errors. Reproduced image data with a small quantization error can be obtained with the amount of compressed image data. Further, in the second embodiment, since the quantization tables Qy and Qc need only be read from the quantization table storage unit 30, quantization can be performed without waiting for the calculation result of the quantization table, and the processing time can be reduced.

[0069]

According to the present invention, the quantization error is predicted,
It is possible to provide an image compression device and a camera that facilitate image compression with little image degradation due to quantization and inverse quantization.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment of a camera according to the present invention.

FIG. 2 is a diagram showing a quantization table for luminance data and a quantization table for color difference data recommended by JPEG.

FIG. 3 shows a DCT coefficient F _vu of luminance data Y and a quantized DC
It is a figure which shows the T coefficient _Rvu , the inverse quantization DCT coefficient _F'vu , and the quantization table Qy.

FIG. 4 shows a DCT coefficient F _vu and an inversely quantized DCT coefficient F ′ _vu
FIG. 9 is a diagram showing a coefficient value error i _vu for each coefficient.

FIG. 5 is a block diagram illustrating a flow in a quantization table creation circuit.

FIG. 6 is a table showing a correspondence between a quantization coefficient q and an rms value.

FIG. 7 shows a created quantization table for luminance data Qya.
FIG. 7 is a diagram showing a color difference data quantization table Qca.

FIG. 8 is a diagram illustrating a first half of a flowchart of a quantization error statistical process;

FIG. 9 is a diagram illustrating the latter half of the flowchart of the quantization error statistical process.

FIG. 10 is a diagram illustrating a first half of a flowchart of a quantization error prediction process.

FIG. 11 is a diagram illustrating the latter half of the flowchart of the quantization error prediction process.

FIG. 12 is a block diagram showing a second embodiment of the camera according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Compressed image apparatus 12 DCT processing circuit 14 Quantization processing circuit 20 Quantization table preparation circuit 21 Statistic amount storage part 22 Quantization error prediction part 24 Quantization table calculation part

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] November 19, 1997

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0022

[Correction method] Change

[Correction contents]

Using the quantization table Qy , the DCT coefficient F
_The equation for quantizing _vu is defined by equation (1). Round in this equation means approximation to the nearest integer.
In other words, the DCT coefficient F _vu and the division of each element of the quantization table Qy , and the rounding-off, make the quantization DC
The T coefficient R _vu is determined.

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0023

[Correction method] Change

[Correction contents]

R _vu = round (F _vu / q _vu ) (1) {0 ≦ u, v ≦ 7}

[Procedure amendment 3]

[Document name to be amended] Statement

[Correction target item name] 0025

[Correction method] Change

[Correction contents]

In order to decompress the coded compressed image signal and display it on the screen, decoding, inverse quantization, two-dimensional DC
Processing of the inverse transformation of T (hereinafter referred to as two-dimensional IDCT) is required. This decoding is the reverse of the operation of Huffman coding, and is well known in the art and will not be described in detail. The quantized DCT coefficients obtained by the decoding are inversely quantized using the quantization tables Qy and Qc used for quantization, and are converted into inversely quantized DCT coefficients F ′ _vu . These inversely quantized DCT coefficients F ′ _vu are second-order inverse transforms of two-dimensional DCT.
The original IDCT is performed, and the data is converted into luminance data Y ′ and chrominance data Cb ′ and Cr ′, respectively. Since the two-dimensional IDCT is also known, it will not be described in detail here.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0042

[Correction method] Change

[Correction contents]

[0042]

(Equation 2)

[Procedure amendment 5]

[Document name to be amended] Statement

[Correction target item name] Explanation of sign

[Correction method] Change

[Correction contents]

[Description of Signs] 10 Image compression apparatus 12 DCT processing circuit 14 Quantization processing circuit 20 Quantization table creation circuit 21 Statistic amount storage unit 22 Quantization error prediction unit 24 Quantization table calculation unit

Claims (9)

[Claims] 1. A quantization means for quantizing an orthogonal transform coefficient corresponding to original image data obtained from an imaging optical system using a quantization table to obtain a quantized orthogonal transform coefficient, based on the orthogonal transform coefficient. And a quantization table creating means for updating the quantization table. 2. The method according to claim 1, wherein said quantization table creating means is configured to: inversely quantize said quantized orthogonal transform coefficient and obtain an orthogonal quantization coefficient whose absolute value is greater than 2; A storage unit for counting the number of samples of transform coefficients for each of the orthogonal transform coefficients, and storing data of the number of samples (sample number data) in each image for a predetermined number of images. 2. The image compression device according to 1. 3. The storage unit deletes the sample number data of the oldest image stored in the storage unit every time the sample number data of a new image is input, thereby obtaining the predetermined number of samples. The image compression apparatus according to claim 2, wherein the sample number data is updated. 4. The quantization table creation means includes: a quantization error prediction unit that predicts the error based on a predetermined number of sample number data; and a prediction error value obtained from the quantization error prediction unit. The image compression apparatus according to claim 3, further comprising: a quantization table calculation unit that calculates a quantization table based on the quantization table and outputs the quantization table to the quantization unit. 5. The quantization table creating means creates the quantization table from the predetermined number of sample number data including the newly input sample number data. An image compression device according to claim 1. 6. The quantization table creating means creates the quantization table based on the predetermined number of sample number data input before the newly input sample number data. The image compression device according to claim 4. 7. A quantization means for quantizing an orthogonal transform coefficient corresponding to original image data obtained from an imaging optical system by using a quantization table to obtain a quantized orthogonal transform coefficient, based on the orthogonal transform coefficient. And a quantization table creating means for updating the quantization table, wherein the quantization table creating means corresponds to image data of one image, and is obtained by inversely quantizing the quantized orthogonal transform coefficients. The number of samples of the orthogonal transform coefficient whose absolute value of the error between the quantized orthogonal transform coefficient and the orthogonal transform coefficient is larger than 2 is counted for each of the orthogonal transform coefficients, and data of the sample number in each image (sample number) Data) for storing a predetermined number of images for a predetermined number of images, and predicting the error based on a predetermined number of the sample number data stored in the sample number data storage means. And a quantization table calculation unit that calculates a quantization table based on the prediction error value obtained from the quantization error prediction unit and outputs the quantization table to the quantization unit. Recording means for recording, on a recording medium, compressed image data image-compressed using the quantization table created by the quantization table creation means, and quantization table storage means for storing the created quantization table. A camera characterized in that: 8. The recording device according to claim 1, wherein the compressed image data is recorded in a first area of the recording medium, and a quantization table used for quantization is recorded in a second area of the recording medium. The camera according to claim 7, characterized in that: 9. The quantization table creation unit uses the quantization table stored in the table storage unit for quantization, and after the recording of the compressed image data on the recording medium is completed, Updating the quantization table by creating the quantization table based on the sample number data corresponding to the compressed image data and storing the created quantization table in the quantization table storage means. 9. The camera according to claim 8, wherein: