JP2002016922A - Image encoder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image encoder capable of obtaining an image code high in quality and stability by controlling the amount of the generated code so as not to cause a change in quantization accuracy. SOLUTION: A frame memory A 1 rearranges an input image frame. An orthogonal converter 2 orthogonally converts either an input image or a difference between a predictive image and the input image. An orthogonal conversion coefficient separator 10 divides the orthogonal conversion coefficient calculated by the orthogonal converter 2 into an encoding coefficient and a non-encoding coefficient. A distortion volume calculator 11 calculates the distortion volume of an image that is induced by cutting off the non-encoding coefficient. An orthogonal conversion coefficient corrector 12 corrects the encoding coefficient so as to lessen the distortion volume. An encoding coefficient output 13 outputs either an encoding coefficient separated by the orthogonal conversion coefficient separator 10 or an encoding coefficient corrected by the orthogonal conversion coefficient corrector 12 to a quantization part 3 to select one of them.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image encoding apparatus for compressing and encoding digital image data.

[0002]

2. Description of the Related Art FIG.
49 is a conventional image encoding device shown in NO.4, in which 1 is a frame memory A for temporarily storing an input image, and 2 is a discrete cosine image for the image stored in the frame memory A1. Transform (DCT: Discrete Cosine Transfo
rm), 3 is a quantization unit for quantizing the orthogonal transformation coefficient calculated by the orthogonal transformation unit 2, 4 is quantization data output from the quantization unit 3, and other codes. Variable-length encoding of encoded data in a predetermined format (hereinafter, referred to as
VLC unit for performing variable length coding (VLC), 5 a buffer for temporarily storing the output of the VLC unit 4, and 6 dequantizing the quantized data output from the quantizing unit 3 and outputting an orthogonal transform coefficient. 7 is an inverse orthogonal transform unit that performs inverse orthogonal transform on the orthogonal transform coefficient output from the inverse quantizer 6, and 8 is an image obtained as an output of the inverse orthogonal transform unit 7, or an inverse orthogonal transform. Frame memories B and 9 for temporarily storing an image obtained by adding image difference data described later obtained as an output of the conversion unit 7 and a prediction image used for generating the difference data are images stored in the frame memory B 8 A motion compensating unit that detects a motion between the image and a newly encoded image as a motion vector.

Next, the operation will be described. The image input from the outside of the image encoding device is temporarily stored in the frame memory A.
1 is stored. For example, ISO / IEC 138
When using 18-2, the display order and the encoding order of the image frames may be different.
A1 rearranges frames. Next, one frame (an image for one screen) of the image stored in the frame memory A 1 is extracted and input to the orthogonal transformation unit 2.
The orthogonal transform unit 2 performs an orthogonal transform on either the input image itself or difference data between a predicted image obtained by performing motion compensation on the image stored in the frame memory B8 and the input image. The quantization unit 3 performs quantization on the orthogonal transform coefficient output from the orthogonal transform unit 2 at an appropriate quantization step. The quantization step can be changed for each frame or macroblock. The orthogonal transform coefficients quantized by the quantization unit 3 are variable-length coded by the VLC unit 4 together with other coded information. The image data variable-length coded by the VLC unit 4 is temporarily stored in the buffer 5 as a coded bit string, and then output to the outside of the image coding apparatus.

[0004] The quantized orthogonal transform coefficients output from the quantization unit 3 are also input to an inverse quantization unit 6. The inverse quantization unit 6 inversely quantizes the orthogonal transform coefficients in a quantization step equivalent to that of the quantization unit 2. The inverse orthogonal transform unit 7 performs an inverse orthogonal transform on the orthogonal transform coefficient inversely quantized by the inverse quantizer 6. The image obtained by the inverse orthogonal transform unit is stored in the frame memory B8.

[0005] The image stored in the frame memory B 8 is equivalent to a decoded coded bit string obtained by a series of processes of the orthogonal transform unit 2, the quantization unit 3, and the VLC unit 4.
It does not always match the original image. This is because an error of the orthogonal transform coefficient generated in the processing of the quantization unit 3 causes image distortion. In general, the degree of distortion tends to increase as the quantization step increases.

By the way, the orthogonal transformation unit 2, the quantization unit 3, the VL
The code amount (hereinafter, generated code amount) of a coded bit string generated by a series of processes of the C unit 4 changes depending on the content of an image and coding parameters. The buffer 5 exists for the purpose of absorbing the difference between the generated code amount and the output bit rate. However, if the difference is large, the buffer 5 underflows or overflows and cannot output a normal encoded bit string. Therefore, means for controlling the generated code amount by increasing or decreasing the quantization step size according to the accumulation amount of the buffer 5 in the quantization unit 3 is used.

[0007]

The conventional image coding apparatus can control the generated code amount by increasing or decreasing the quantization step size in the quantization unit 3, but the quantization is performed in units of macroblocks or the like. If the steps change drastically,
There has been a problem that the image quality is extremely deteriorated due to a variation in the quantization accuracy, and the like. In particular, when the quantization accuracy of the low-frequency component of the orthogonal transform coefficient changes drastically, there is a possibility that very severe image quality deterioration may occur.

The present invention has been made to solve the above problems, and realizes high image quality and high stability by controlling the generated code amount so as not to cause a change in quantization accuracy. It is an object of the present invention to provide an image encoding device that performs the following.

[0009]

An image encoding apparatus according to the present invention comprises a first frame memory for rearranging the encoding order of input image frames, and an image frame stored in the first frame memory. An orthogonal transform unit that performs an orthogonal transform such as a discrete cosine transform (DCT), a quantizing unit that quantizes orthogonal transform coefficients calculated by the orthogonal transform unit and outputs quantized data, A VLC unit that performs variable length coding (hereinafter, VLC: Variable Length Coding) of the quantized data in a predetermined format, a buffer for temporarily storing the output of the VLC unit, and a buffer output from the quantizer. An inverse quantization unit that inversely quantizes the quantized data to output an orthogonal transformation coefficient, an inverse orthogonal transformation unit that performs an inverse orthogonal transformation on the orthogonal transformation coefficient output from the inverse quantization unit, Inverse orthogonal transform A second frame memory for temporarily storing image data obtained as a force, and a motion for detecting a motion between an image stored in the second frame memory and a newly encoded image as a motion vector And a compensating unit, wherein the image coding apparatus compresses and encodes the input image frame by applying an orthogonal transform to the input image frame, among the orthogonal transform coefficients output from the orthogonal transform unit, orthogonal transform coefficients of a specific frequency component Is provided with an orthogonal transform coefficient separation unit that cuts off the image without encoding, and a distortion amount calculation unit that calculates the amount of image distortion caused by cutting off the orthogonal transform coefficient.

[0010] The distortion amount calculator performs inverse orthogonal transform on the orthogonal transform coefficient to be cut off.

Further, the distortion amount calculating section makes a prediction from an accumulated value of orthogonal transform coefficients to be cut off or the like.

The orthogonal transform coefficient separating section determines a frequency component to be cut off so that the amount of image distortion caused by cutting off the orthogonal transform coefficient does not increase.

The orthogonal transform coefficient separating means changes a frequency component to be cut off in accordance with the amount of data stored in the buffer.

The orthogonal transform coefficient separating means increases the frequency components to be cut off as the buffer storage amount increases.

Further, of the orthogonal transform coefficients to be cut off and the orthogonal transform coefficients not to be cut off, which are output by the orthogonal transform coefficient separating section by separating the orthogonal transform coefficients, the orthogonal transform coefficients to be not cut off are selected by using the orthogonal transform coefficients to be cut off. It has an orthogonal transform coefficient correction unit for correcting.

The orthogonal transform coefficient correction unit includes an inverse orthogonal transform unit for performing an inverse orthogonal transform on an orthogonal transform coefficient to be cut off, a signal transforming unit for transforming an image signal obtained by the inverse orthogonal transform, and an orthogonal transform unit for transforming the transformed signal. And an orthogonal transformation unit that corrects an orthogonal transformation coefficient that is not cut off using the orthogonal transformation coefficient obtained by the orthogonal transformation.

Further, the signal transformation means uses a filter.

In addition, the image processing apparatus further includes a visual characteristics information processing unit that uses information on visual characteristics in determining a frequency component to be cut off.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIG. 1 is a block diagram showing Embodiment 1 of an image coding apparatus according to the present invention. In the figure, 1 is a frame memory A for temporarily storing an input image, and 2 is a discrete cosine transform (DCT) for the image of the frame memory A 1.
rm), 3 is a quantizer for quantizing orthogonal transform coefficients obtained by the orthogonal transform by the orthogonal transformer 2, 4 is quantized data output from the quantizer 3, and others. VLC that performs variable length coding (hereinafter, VLC: Variable Length Coding) of the coded information in a predetermined format
5, a buffer for temporarily storing the output of the VLC unit 4, 6 a dequantizing unit for dequantizing the quantized data output from the quantizing unit 3, and 7 a dequantizing unit An inverse orthogonal transform unit that performs inverse orthogonal transform on the output orthogonal transform coefficient, 8 is a frame memory B for temporarily storing an image obtained as an output of the inverse orthogonal transform unit 7, and 9 is a frame memory B
8 is a motion compensating unit that detects a motion between an image stored in 8 and an image newly input to the orthogonal transformation unit 2 as a motion vector. Reference numeral 10 denotes a coefficient of a frequency component to be encoded (hereinafter, referred to as an encoding coefficient) and a coefficient of a frequency component to be cut off without encoding (hereinafter, referred to as an unencoded coefficient). The orthogonal transform coefficient separating unit 11 for separating the non-coding coefficients is used to calculate the magnitude of image distortion caused by cutting off the non-coding coefficients (hereinafter referred to as “distortion amount”). An orthogonal transform coefficient correction unit 13 corrects the orthogonal transform coefficients for the purpose of performing the coding. An encoding coefficient output unit 13 outputs the orthogonal transform coefficients to be encoded to the quantization unit 3.

Next, the operation will be described. Image data input from outside the image encoding device is temporarily stored in the frame memory A1. For example, ISO / I as an encoding method
When EC 13818-2 is used, the display order of the image frames and the encoding order may be different, so the frames are rearranged in the frame memory A1. Next, one frame of the image stored in the frame memory A 1 is taken out and input to the orthogonal transformation unit 2. Orthogonal transformation unit 2
Is the input image itself or the frame memory B
The orthogonal transformation is performed on any of the difference data between the predicted image obtained by interpolating the image stored in 8 using the motion vector and the input image.

Next, the orthogonal transform coefficient separating section 10 separates the orthogonal transform coefficients calculated by the orthogonal transform section 2 into coded coefficients to be coded and non-coded coefficients not to be coded.
As the non-coding coefficient, the frequency component of the non-coding coefficient is determined so that the coefficient of the high-frequency component is selected as much as possible. The buffer 5 determines which frequency component coefficient is used as the non-coding coefficient.
It is desirable to be able to change it according to the accumulated amount of data. For example, as the buffer storage amount increases as shown in FIG. 2, by increasing the frequency components used as the non-coding coefficients, unnecessary image quality degradation can be prevented.

The distortion calculator 11 calculates the amount of distortion of the image caused by cutting off the non-coding coefficients. As a calculation method, there are a method of performing an inverse orthogonal transform on an uncoded coefficient, a method of performing prediction from an accumulated value of the uncoded coefficient, and the like. These methods are effective because the amount of distortion can be accurately and easily obtained.

The order of the processing performed by the orthogonal transform coefficient separating unit 10 and the distortion amount calculating unit 11 may be reversed. For example, a method of selecting a frequency component to be an uncoded coefficient is f0
(F0> 0) or more, f1 (f1> f0) or more, f2 (f
2> f1) or more, f3 (f3> f2) or more.
After calculating the distortion for each method, the orthogonal transform coefficient separation unit 10 can also separate the coded coefficients and the non-coded coefficients so that the distortion is minimized.

The orthogonal transform coefficient correction unit 12 corrects the coefficient by using a low-pass filter or the like so that the distortion caused by cutting off the non-coded coefficient by correcting the value of the coded coefficient is reduced. Do. The corrected coefficients are output to the coding coefficient output unit 13. Coding coefficient output unit 13
Selects one of the coded coefficients separated by the orthogonal transform coefficient separating unit 10 and the coded coefficients corrected by the orthogonal transform coefficient correcting unit 12, and outputs the selected coefficient to the quantizing unit 3.

The quantization unit 3 performs quantization by selecting either the orthogonal transform coefficient output from the coded coefficient output unit 13 or the orthogonal transform coefficient output from the orthogonal transform unit 2. In this case, a condition such as an allowable time is provided in advance in the quantization unit 3, and, for example, the time required from the orthogonal transformation coefficient separation by the orthogonal transformation coefficient separation unit 21 to the correction of the orthogonal transformation coefficient by the orthogonal transformation coefficient correction unit 27. If the time is within the allowable time, the orthogonal transform coefficient output from the encoding coefficient output unit 13 is selected, and the orthogonal transform coefficient separation by the orthogonal transform coefficient separation unit 21 and the orthogonal transform coefficient If the time required for the correction exceeds the allowable time, the orthogonal transform coefficient output from the orthogonal transform unit 2 is selected. The result of the distortion amount calculator 11 can also be used. During quantization, the quantization step may be changed as usual.

The VLC unit 4 performs variable-length encoding of encoded data such as a motion vector detected by the motion compensating unit 9 in addition to the quantization information output by the quantizing unit 3 (this signal line is not shown). The generated encoded bit string is output to the buffer 5. After temporarily storing the coded bit sequence, the buffer 5 outputs the coded bit sequence to the outside of the image coding device at a predetermined bit rate.

The quantized orthogonal transform coefficients output from the quantization unit 3 are also input to the inverse quantization unit 6. The inverse quantization unit 6 performs inverse quantization at the same quantization step as the quantization unit 3. Next, the quantized orthogonal transform coefficient is subjected to inverse orthogonal transform by the inverse orthogonal transform unit 7. The image obtained by the inverse orthogonal transform is stored in the frame memory B8. The image stored in the frame memory B8 is used to generate a predicted image of an image to be encoded later.

FIG. 3 shows the orthogonal transform coefficient separating unit 10 shown in FIG.
FIG. 3 is a block diagram illustrating an example of a series of encoding coefficient correction processes performed by a distortion amount calculation unit 11, an orthogonal transformation coefficient correction unit 12, and an encoding coefficient output unit 13. First, the orthogonal transform coefficient separating unit 21 separates the orthogonal transform coefficients output from the orthogonal transform unit 2 into coded coefficients to be coded and uncoded coefficients not to be coded. Next, the coded coefficient zeroing means 22 inputs both the coded coefficients and the non-coded coefficients separated from the orthogonal transform coefficient separating means 21 and performs zero substitution of the coded coefficients. For example, in a block of 8 × 8 pixels, a coefficient of a frequency component corresponding to a coding coefficient is replaced with 0. Therefore,
Only the remaining uncoded coefficients are output from the coded coefficient zeroing means 22. Similarly, the non-coding coefficient zeroing means 23
Inputs both the coded coefficient and the non-coded coefficient separated from the orthogonal transform coefficient separating unit 21, and replaces the coefficient of the frequency component corresponding to the non-coded coefficient with 0. Therefore, only the remaining coded coefficients are output from the non-coded coefficient zeroing means 23. Next, the inverse orthogonal transform unit 24 performs an inverse orthogonal transform on the non-coded coefficients output from the coded coefficient zeroing unit 22. As a result, image distortion caused by cutting off the non-coding coefficient is obtained as a discrete signal. The distortion amount calculating means 25 analyzes the obtained discrete signal and calculates the magnitude of the distortion. The orthogonal transform coefficient separating means 21 may separate the number of orthogonal transforms again according to the magnitude of the distortion.
Thereby, the image coding apparatus can know the optimal frequency component to be cut off and the distortion amount to be corrected at that time.

Next, the discrete signal of the distortion obtained by the inverse orthogonal transformation means 24 is input to the signal transformation means 26. The signal transformation means 26 transforms the discrete signal into an approximate signal. The orthogonal transform means 27 performs an orthogonal transform on the approximate signal.

FIG. 4 is an explanatory diagram showing a method of signal transformation by the signal transformation means 26. The components are actually shown in two dimensions, but are shown here in one dimension for simplicity. When performing signal deformation, the signal deformation means 26 uses, for example, a low-pass filter. Note that it is necessary to select a low-pass filter that does not cut high-frequency components but that smoothes high-frequency components and converts them into lower-frequency components. As a result of the signal transformation by the signal transformation unit 26, the orthogonal transformation unit 27 outputs a high-frequency component, which is an uncoded coefficient, from the orthogonal transformation coefficients shown in FIG.
As shown in (b), it is output in a form converted to a low frequency component. In this case, in the orthogonal transform coefficient separating unit 10, the non-coding coefficient zeroing means 23 outputs a coded coefficient which is a low frequency component among the orthogonal transform coefficients shown in FIG. The orthogonal transform coefficient output from the orthogonal transform means 27 shown in FIG. 4B is replaced with the coded coefficient output from the non-coding coefficient zeroing means 23 (lower of the orthogonal transform coefficients shown in FIG. As shown in FIG. 4 (c), the correction is performed by adding the approximate coefficient component to the non-coded coefficient instead of cutting off the non-coded coefficient. The coded coefficient output means 28 outputs either the coded coefficient corrected by addition as shown in FIG. 4C or the coded coefficient output from the non-coded coefficient zeroing means 23. One of them is selected and output to the quantization unit 3 in FIG. By using a low-pass filter in this way, it is easy to realize an image encoding device capable of correcting screen distortion due to clipping.

As described above, according to the first embodiment, since the means for cutting off a specific frequency component from the orthogonal transform coefficients is provided, it is possible to control the generated code amount without greatly changing the quantization step. In addition, it is possible to suppress image quality deterioration due to variation in quantization accuracy. Also, the amount of image distortion due to cutting off the orthogonal transform coefficient is calculated,
By feeding back the result of the calculation to the orthogonal transform coefficient separating means, the frequency component of the orthogonal transform coefficient to be cut off can be adjusted so as to minimize the amount of image distortion, and excessive image quality deterioration can be prevented. In addition, by correcting the value of the orthogonal transform coefficient using the result of calculating the distortion amount, it is possible to reduce the image quality deterioration caused by cutting off the orthogonal transform coefficient.

When a filter is used as the signal transforming means 26, it is desirable to select the type of the filter in consideration of how to separate the frequency components of the coded coefficients and the non-coded coefficients and the characteristics of the orthogonal transform.

Embodiment 2 FIG. When information on human visual characteristics can be obtained, if the information is used to adjust the frequency components to be separated by the orthogonal transform coefficient separation means 21, image quality degradation can be reduced. The first embodiment does not include a means for handling information on visual characteristics, but the second embodiment includes a visual characteristics information processing unit. FIG. 5 is a block diagram showing Embodiment 1 of the image encoding apparatus of the present invention. 5, the same reference numerals as those in FIG. 1 denote the same or corresponding parts. The visual characteristic information processing unit 14 calculates, for example, values and distributions of luminance and color differences contained in an image, averages and variances of luminance and color differences in a specific area, and analyzes visual characteristics based on the calculated values. is there. By using these pieces of information, for example, as shown in FIG. 6, the frequency component to be used as a non-coding coefficient is narrowed in a region where deterioration is conspicuous visually, and the frequency component to be used as a non-coding coefficient is specified in a region where deterioration is not conspicuous. Processing such as widening becomes possible.

As described above, according to the second embodiment, in addition to the effect of the first embodiment, by using information on visual characteristics, excessive image quality deterioration can be prevented and efficient image quality control can be performed. Become.

【The invention's effect】

As described above, according to the present invention,
Among the orthogonal transform coefficients output from the orthogonal transform unit, an orthogonal transform coefficient separating unit that cuts off an orthogonal transform coefficient of a specific frequency component without encoding, and calculates an image distortion amount caused by cutting off the orthogonal transform coefficient. With the provision of the distortion amount calculating section, the amount of generated codes can be controlled while suppressing image distortion.

Further, since the distortion amount calculation unit performs inverse orthogonal transformation on the orthogonal transformation coefficient to be cut off, there is an effect that the distortion amount can be accurately and easily obtained.

Further, since the distortion amount calculation unit makes predictions based on the accumulated value of the orthogonal transform coefficients to be cut off, it is possible to obtain the amount of distortion accurately and easily.

Further, since the orthogonal transform coefficient separating section determines the frequency component to be cut off so as not to increase the distortion amount of the image caused by cutting off the orthogonal transform coefficient, high image quality can be maintained. This has the effect.

Further, since the orthogonal transform coefficient separating means changes the frequency to be cut off in accordance with the accumulation amount of the buffer, there is an effect that unnecessary image quality deterioration can be prevented.

Further, since the orthogonal transform coefficient separating means increases the frequency components to be cut off as the buffer storage amount increases, there is an effect that unnecessary image quality deterioration can be prevented.

Further, of the orthogonal transform coefficients to be cut off and the orthogonal transform coefficients not to be cut off, which are output by the orthogonal transform coefficient separating section by separating the orthogonal transform coefficients, the orthogonal transform coefficients to be not cut off are selected by using the orthogonal transform coefficients to be cut off. The provision of the orthogonal transform coefficient correction unit for performing the correction makes it possible to alleviate the image quality deterioration due to the cutoff of the orthogonal transform coefficient.

The orthogonal transform coefficient correction unit includes an inverse orthogonal transform unit for performing an inverse orthogonal transform on an orthogonal transform coefficient to be cut off, a signal transforming unit for transforming an image signal obtained by the inverse orthogonal transform, and an orthogonal transform unit for transforming the transformed signal. Since the orthogonal transformation coefficient is corrected by using the orthogonal transformation coefficient obtained by the orthogonal transformation, there is an effect that the suppression of the image quality deterioration due to the cutoff of the orthogonal transformation coefficient can be efficiently realized. .

Further, since the signal transformation means uses a filter, it is easy to realize an image coding apparatus capable of correcting screen distortion due to partial cut-off of orthogonal transform coefficients.

In determining the frequency components to be cut off, information on visual characteristics is used.
This has the effect of preventing excessive image quality deterioration and enabling efficient image quality control.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating Embodiment 1 of an image encoding device according to the present invention.

FIG. 2 is an example of separation of coded coefficients and uncoded coefficients.

3 is a block diagram showing an embodiment of a series of encoding coefficient correction processes performed by an orthogonal transformation coefficient separation unit 10, a distortion amount calculation unit 11, an orthogonal transformation coefficient correction unit 12, and an encoding coefficient output unit 13 in FIG. It is.

FIG. 4 is an explanatory diagram showing a method of orthogonal transform coefficient correction.

FIG. 5 is a block diagram illustrating Embodiment 2 of an image encoding device according to the present invention.

FIG. 6 is a diagram illustrating a relationship between a buffer accumulation amount and a conspicuous change in image quality.

FIG. 7 is a block diagram illustrating a conventional image encoding device.

[Explanation of symbols]

 Reference Signs List 1 frame memory A 2 orthogonal transformation unit 3 quantization unit 4 VLC unit 5 buffer 6 inverse quantization unit 7 inverse orthogonal transformation unit 8 frame memory B 9 motion compensation unit 10 orthogonal transformation coefficient separation unit 11 distortion calculation unit 12 orthogonal transformation coefficient Correction unit 13 Coding coefficient output unit 14 Visual characteristic information processing unit 21 Orthogonal transformation coefficient separation unit 22 Coding coefficient zeroization unit 23 Non-coding coefficient zeroization unit 24 Inverse orthogonal transformation unit 25 Distortion amount calculation unit 26 Signal deformation unit 27 Orthogonal transformation means 28 Coding coefficient output means.

Claims (10)

[Claims] 1. A first frame memory for rearranging the coding order of input image frames, and a discrete cosine transform (DCT) for image frames stored in the first frame memory. An orthogonal transform unit that performs orthogonal transform of the above, a quantizer that quantizes the orthogonal transform coefficient calculated by the orthogonal transform unit, and outputs quantized data, and a variable-length encoding of the quantized data in a predetermined format ( Hereinafter, a VLC unit for performing Variable Length Coding (VLC), a buffer for temporarily storing the output of the VLC unit, and an orthogonal transform coefficient obtained by dequantizing the quantized data output from the quantization unit. An inverse quantizer for outputting, an inverse orthogonal transformer for performing inverse orthogonal transform on the orthogonal transform coefficient output from the inverse quantizer, and temporarily storing image data obtained as an output of the inverse orthogonal transformer. The first to accumulate A second frame memory, and a motion compensator for detecting a motion between the image stored in the second frame memory and the newly encoded image as a motion vector. In the image encoding apparatus that performs compression encoding by applying, among the orthogonal transform coefficients output from the orthogonal transform unit, an orthogonal transform coefficient separating unit that cuts off without encoding the orthogonal transform coefficient of a specific frequency component. Image encoding apparatus, comprising: a distortion amount calculation unit that calculates an image distortion amount caused by cutting off orthogonal transform coefficients. 2. The distortion calculation unit according to claim 1, wherein the orthogonal transform coefficient to be cut off is subjected to inverse orthogonal transform.
The image encoding device according to claim 1. 3. The image encoding apparatus according to claim 1, wherein the distortion amount calculating unit performs prediction based on an accumulated value of orthogonal transform coefficients to be cut off. 4. The method according to claim 1, wherein the orthogonal transform coefficient separating unit determines a frequency component to be cut off so that an amount of distortion of an image caused by cutting off the orthogonal transform coefficient does not increase. Image encoding device. 5. The image encoding apparatus according to claim 1, wherein the orthogonal transform coefficient separating unit changes a frequency component to be cut off in accordance with a storage amount of the buffer. 6. The image encoding apparatus according to claim 5, wherein the orthogonal transform coefficient separating unit increases the frequency components to be cut off as the buffer storage amount increases. 7. An orthogonal transform coefficient to be cut off, which is output by an orthogonal transform coefficient separation unit separating and outputting an orthogonal transform coefficient, and an orthogonal transform coefficient that is not cut off using the orthogonal transform coefficient to be cut off, among the orthogonal transform coefficients to be cut off. The image coding apparatus according to claim 1, further comprising an orthogonal transform coefficient correction unit that corrects the image. 8. An orthogonal transform coefficient correction unit, comprising: an inverse orthogonal transform unit for performing an inverse orthogonal transform on an orthogonal transform coefficient to be cut off; a signal transforming unit for transforming an image signal obtained by the inverse orthogonal transform; 8. The image coding apparatus according to claim 7, further comprising an orthogonal transform unit that corrects an orthogonal transform coefficient that is not cut off using the orthogonal transform coefficient obtained by the orthogonal transform. 9. The image encoding apparatus according to claim 8, wherein the signal transformation unit uses a filter. 10. The image encoding apparatus according to claim 1, further comprising a visual characteristic information processing unit that uses information relating to visual characteristics in determining a frequency component to be cut off.