JP2002077912A - Image encoder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image encoder in which a plurality of image data compressed at a plurality of compression rates can be outputted in parallel from a plurality of output sections. SOLUTION: A discrete cosine transform unit Q1 transforms original image data of 8×8 pixels into DCT coefficients of 8×8 arrangement. A first operating section Q2 outputs convenient DCT coefficients where 5×5 elements arranged on the low frequency side are values transformed from the original image data and 0 is set for each of remaining elements on the high frequency side. Second operating sections Q1-Q2 output DCT coefficients where all elements of 8×8 arrangement are set with values transformed from the original image data. A switch S3 is interlocked with a switch S1 and only a convenient DCT count only subjected to in-frame compression is inputted to a second quantizer SMP 2. A first quantizer SMP 1 quantizes data to 8 bit values and the second quantizer SMP 2 quantizes data to 7 bit values.

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 generating compressed image data from input original image data and outputting the compressed image data.

[0002]

2. Description of the Related Art Conventionally, original image data is input, and the original image data is converted into compressed image data having a smaller data amount by encoding the original image data in accordance with an image compression encoding system represented by, for example, the MPEG standard. An image encoding device that outputs the compressed image data is known.

[0003] This type of image encoding apparatus is used, for example, when transmitting a moving image via a network network represented by the Internet or a digital broadcasting network represented by HDTV, or a recording apparatus capable of recording digital signals. It is used when recording a moving image on a medium.

[0004]

In the above-mentioned networks and digital broadcasting networks, the transmission bandwidth and transmission characteristics of each transmission line are different. Therefore, compressed image data transmitted through each transmission line is usually Are compressed at a compression ratio suitable for each transmission path. Also, when the compressed image data is recorded on the recording medium, the optimal compression ratio is selected in consideration of the image quality, the recording capacity of the recording medium, and the like.

[0005] Therefore, for example, when compressed image data is simultaneously transmitted to both a network and a digital broadcasting network, or when compressed image data is simultaneously transmitted to a network and compressed image data is recorded on a recording medium,
Conventionally, when simultaneously outputting compressed image data to a plurality of different transmission paths and output destinations, a plurality of image encoding devices corresponding to a plurality of different transmission paths and output destinations are prepared, and each transmission path and Compressed image data having an optimal compression ratio according to the characteristics of the output destination is output from each image encoding device.

However, such a method requires a plurality of image encoding devices corresponding to a plurality of different transmission paths and output destinations, and therefore costs are increased accordingly.
There was a disadvantage. Even without using a plurality of image encoding devices, a plurality of compressed image data corresponding to each of a plurality of different transmission paths and output destinations may be created in advance using one image encoding device. In other words, after the plurality of compressed image data have been prepared, each compressed image data can be simultaneously transmitted to a plurality of transmission paths and output destinations.

[0007] However, such a technique can only cope with a case where compressed image data compressed at a plurality of compression ratios can be prepared in advance, and a plurality of compressed image data can be obtained by real-time processing. Therefore, for example, simultaneous live broadcasting using both the network network and the digital broadcasting network or recording on a recording medium simultaneously with such live broadcasting cannot be performed.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problem, and an object of the present invention is to provide a device configuration which is less costly than using a plurality of image encoding devices, and which is compressed at a plurality of compression ratios. It is another object of the present invention to provide an image encoding apparatus capable of outputting a plurality of compressed image data from a plurality of output units in parallel.

[0009]

According to the first aspect of the present invention, there is provided an image encoding apparatus configured to achieve the above object, by converting original image data input from a single input unit into Image encoding that generates a plurality of compressed image data compressed at a plurality of compression ratios by performing encoding by dividing into a plurality of encoding processing systems, and outputting each compressed image data from a plurality of output units in parallel An apparatus, wherein at least a first encoding processing system encodes the original image data to generate first compressed image data having a data amount smaller than the original image data; An encoding processing system selectively extracts a predetermined part of data from the intermediate data partially processed by the first encoding processing system, based on the extracted part of data, The data amount is the first And generates a second compressed image data to be smaller than the reduced image data.

In this image encoding apparatus, the first and second
The first encoding processing system encodes original image data input from the same input unit, the first encoding processing system generates first compressed image data, and the second encoding processing system Generates second compressed image data that is smaller than the first compressed image data.

Therefore, according to this image encoding apparatus, compressed image data having different compression rates are output in parallel to two or more different transmission paths and output destinations without preparing a plurality of image encoding apparatuses. be able to. In particular, in this image encoding apparatus, the second encoding processing system selectively extracts a predetermined part of data from the intermediate data partially processed by the first encoding processing system. And generating second compressed image data based on the extracted partial data. For this reason, the data processing means required from the input of the original image data to the generation of the intermediate data is shared by the first and second encoding processing systems.

Therefore, according to this image coding apparatus, the apparatus configuration is more than that of the apparatus having only a plurality of sets of all necessary data processing means from input of original image data to output of compressed image data. Can be simplified, so that a device configuration that does not cost as much as using a plurality of image encoding devices can be achieved.

This image coding apparatus includes a plurality of coding processing systems including at least first and second coding processing systems. However, when the image coding apparatus includes three or more coding processing systems, The third and subsequent encoding processing systems may be configured in any manner. For example, the third and subsequent encoding processing systems may have the same configuration as any of the first and second encoding processing systems, or may have a completely different configuration. Further, similarly to the case where the second encoding processing system uses the intermediate data partially processed by the first encoding processing system, the third and subsequent encoding processing systems use the first encoding processing system or A configuration may be used in which intermediate data that has been partially processed by the second encoding processing system is used.

Next, the configuration of the first and second encoding processing systems will be described more specifically. First, in the image encoding apparatus according to claim 2, the first encoding processing system converts the original image data in a two-dimensional array into DCT coefficients in a two-dimensional array by discrete cosine transform, and Generating the first compressed image data based on the coefficient; and the second encoding processing system calculates the first compressed image data from the DCT coefficient converted by the first encoding processing system from a frequency array range of the DCT coefficient. A low-frequency array element in a low-frequency array range is selectively extracted, and all array elements other than the low-frequency array element are set to 0 (zero) to generate a simple DCT coefficient, based on the simple DCT coefficient. The method is characterized in that the second compressed image data is generated.

In this image encoding apparatus, the first encoding processing system converts the original image data of the two-dimensional array into DCT coefficients of the two-dimensional array by discrete cosine transform. On the other hand, the second encoding processing system selectively extracts low-frequency-side array elements from the DCT coefficients transformed by the first encoding processing system, and sets all array elements other than the low-frequency-side array elements to 0. (0) to generate a simple DCT coefficient.

The first encoding processing system generates first compressed image data based on the DCT coefficients, and the second encoding processing system generates second compressed image data based on the simplified DCT coefficients. Generate The method of generating the first and second compressed image data is not limited as long as the second compressed image data is smaller than the first compressed image data. However, for example, a processing procedure adopted in a well-known MPEG standard can be arbitrarily adopted. More specifically, for example, the DC
Processing such as dividing the T coefficient or the simplified DCT coefficient by an arbitrarily selected divisor and quantizing, or scanning the array elements in a predetermined order (for example, zigzag scanning) to form a one-dimensional coefficient sequence, as necessary. Then, a method of compressing the data amount by utilizing the bias of the probability of occurrence of data by entropy coding (variable length coding) can be adopted.
At this time, the simplified DCT coefficient is expected to have more 0 (zero) in the array element compared to the base DCT coefficient.
If a processing procedure such that 0 (zero) is compressed at a higher rate is included, the data amount of the second compressed image data may be smaller than the data amount of the first compressed image data. it can.

According to the image coding apparatus configured as described above, two or more compressed image data having different data amounts (compression ratios) can be output as described above. Compressed image data having different compression rates can be output to an output destination in parallel.

The data processing means required from input of the original image data to conversion of the original image data into DCT coefficients is shared by the first and second encoding processing systems. A simple device configuration that does not cost much can be provided. Next, in the image encoding device according to claim 3, the first encoding processing system quantizes the DCT coefficient, and the first encoding processing system performs quantization on the first coefficient based on the quantized coefficient.
, And converts the quantized coefficients into decoded image data by inverse quantization and inverse discrete cosine transform, and performs inter-frame compression based on the decoded image data and the original image data. The second encoding processing system calculates the simplified DCT coefficient from the DCT coefficients subjected to the inter-frame compression by the first encoding processing system.
It is characterized by generating a T coefficient.

In this image encoding apparatus, the inter-frame compression is performed by taking the difference between one image and the other image by utilizing the high correlation between two images adjacent on the time axis. A compression method for reducing the amount of data by encoding. For example, a P picture and a B picture in the MPEG standard are inter-frame compressed images.

In this image encoding apparatus, the first encoding processing system performs inter-frame compression when generating the first compressed image data. In comparison, the first compressed image data is compressed at a higher rate, and the data amount is reduced.

Further, the second encoding processing system generates the above-described simplified DCT coefficients from the DCT coefficients obtained by performing the inter-frame compression by the first encoding processing system. As compared with the case where the decoded image data is generated in the processing system and the inter-frame compression is performed, the deterioration of the image quality can be suppressed.

More specifically, in the second encoding system, when the simplified DCT coefficient is quantized and the second compressed image data is generated based on the quantized coefficient, the quantized DCT coefficient is quantized. It is possible to convert the coefficients into decoded image data by inverse quantization and inverse discrete cosine transform, and it is also possible to perform inter-frame compression based on the decoded image data and original image data. However, since the amount of data of the simple DCT coefficient is smaller than that of the DCT coefficient on which it is based, the second compressed image data tends to become an image having a low spatial resolution at the time of decoding. Even if DCT coefficients are obtained by performing inter-frame compression based on, the image quality will deteriorate at this point.

On the other hand, the DCT coefficient used in the first encoding processing system has a larger data amount than the simplified DCT coefficient handled in the second encoding processing system, and the first compressed image data is An image having a higher spatial resolution is obtained at the time of decoding as compared with the compressed image data of No. Therefore, if DCT coefficients are obtained by performing inter-frame compression based on the high-resolution decoded image data and the original image data, the degree of image quality deterioration is suppressed at this point.

Therefore, if the simplified DCT coefficients are generated from the DCT coefficients obtained by performing the inter-frame compression by the first encoding processing system, the decoded image data is generated in the second encoding processing system to generate the frame data. It is possible to suppress deterioration of the image quality as compared with the case where the inter-compression is performed.

Next, in the image encoding apparatus according to the fourth aspect, the first encoding processing system generates the first compressed image data subjected to intra-frame compression and inter-frame compression, The second encoding processing system selectively extracts data corresponding to the frame subjected to only the intra-frame compression from the intermediate data partially encoded by the first encoding processing system, The second compressed image data is generated based on the extracted data.

In this image coding apparatus, the intra-frame compression means that one pixel and another pixel are used, for example, by utilizing the high correlation between two adjacent pixels in a single image. A compression method that reduces the amount of image data by encoding the difference.
An I picture referred to in the standard is an image that has been compressed within a frame. Also, the inter-frame compression is, as described above,
A compression method that reduces the amount of data by coding by taking the difference between one image and the other image using the high correlation between two images that are adjacent on the time axis. For example, P pictures and B pictures in the MPEG standard are inter-frame compressed images.

In this image encoding device, the first encoding processing system generates first compressed image data that has been subjected to intra-frame compression and inter-frame compression. On the other hand, the second encoding processing system selectively extracts data corresponding to a frame subjected to only intra-frame compression from the intermediate data partially encoded by the first encoding processing system,
The second compressed image data is generated. Therefore, the second compressed image data is data with a low time resolution obtained by thinning out a part of the first compressed image data, and the data amount is smaller than that of the first compressed image data. Further, since the second compressed image data is constituted by data corresponding to a frame subjected to only intra-frame compression, the second compressed image data corresponds to preceding and succeeding frames even though some frames are thinned out. There is no need for data, and a decoded image can be obtained without any problem.

Next, an image encoding apparatus according to claim 5, wherein the first encoding system divides the intermediate data generated in the course of the encoding process by a first divisor to obtain a quantum value. And performing quantization by dividing the intermediate data generated in the middle of the encoding process by a second divisor larger than the first divisor. And

In this image coding apparatus, the first and second
The encoding processing system divides the intermediate data generated in the middle of the encoding processing by first and second divisors having different sizes, so that the second compressed image data is Compared to the compressed image data, the data amount is smaller and the amplitude resolution is lower.

[0030]

Next, an embodiment of the present invention will be described with reference to an example. As shown in FIG. 1, an image coding apparatus described as an embodiment of the present invention includes a discrete cosine transformer Q1, a first quantizer SMP1, and a second quantizer SM.
P2, first entropy coder C1, second entropy encoder C2, inverse quantizer SMP1 ^-1, the inverse discrete cosine transformer Q1 ^-1, with the prediction memory M, the loop filter F, and switches S1~S3 The original image data is input from a single input unit, and the first and second compressed image data are output in parallel from the first and second output units.

The discrete cosine transformer Q1 is means for transforming the original image data into DCT coefficients by the discrete cosine transform. The discrete cosine transform unit Q1 uses an image composed of pixels of a two-dimensional array (8 × 8 in the present embodiment) as a unit of transformation. Original image data representing an image serving as a unit is converted into DCT coefficients in a two-dimensional array (8 × 8 in the present embodiment) by discrete cosine transform.
In the present embodiment, the discrete cosine transformer Q1 is
It is composed of an operation unit Q2 and a second operation unit Q1-Q2. As shown in FIG. 2, the first calculation unit Q2 converts the 5 × 5 low-frequency array elements of the finally obtained 8 × 8 DCT coefficients from the original image data, and A simple DCT coefficient is generated in which all 0 (zero) are set in the high-frequency side array elements of. Also, the second operation unit Q1-
Q2 sets a value converted from the original image data also for the above-described high-frequency side array element, and generates a DCT coefficient in which all 8 × 8 array elements are converted from the original image data. In the block diagram of FIG. 1, for convenience, the first
After the operation unit Q2 generates the simple DCT coefficient, the second operation unit Q1
Although a system for completing the DCT coefficient by -Q2 is shown, a specific processing procedure for obtaining the DCT coefficient and the simplified DCT coefficient inside the discrete cosine transformer Q1 is arbitrary, for example, an 8 × 8 DCT After generating the coefficients, a part of the DCT coefficients is set to 0 (zero) to
It may be configured to generate a T coefficient.

The first quantizer SMP1 is means for quantizing the DCT coefficient by dividing the DCT coefficient by a first divisor.
The quantizer SMP2 is means for quantizing the simple DCT coefficient by dividing it by the second divisor. In the first divisor and the second divisor, the second divisor is larger, and in the present embodiment, the data value output from the first quantizer SMP1 is quantized to an 8-bit value (0 to 255). The data value output from the second quantizer SMP2 is quantized to a 7-bit value (0 to 127).

The first entropy encoder C1 and the second
The entropy coder C2 is means for performing data conversion by a well-known scheme such as Huffman coding or the like so that data having a higher appearance frequency has a shorter data length according to the occurrence probability of the data, thereby compressing the data.

The inverse quantizer SMP1 ^-1 and the inverse discrete cosine transformer Q1 ^-1 are means for performing decoding in a procedure reverse to the encoding procedure by the discrete cosine transformer Q1 and the first quantizer SMP1. . The prediction memory M includes an inverse quantizer S
This is means for storing the decoded image data decoded by MP1 ^-1 and the inverse discrete cosine transformer Q1 ^-1 .

The loop filter F is a means for reducing noise components in the decoded image data generated during decoding. In the image coding apparatus configured as described above, original image data is input from a single input unit. The original image data has been subjected to necessary pre-processing (for example, frame rearrangement, etc.) in advance according to a predetermined compression method such as the MPEG standard. In this embodiment, 8 × 8 pixels are used. The original image data is input using the image as a conversion unit.

The original image data input from the input unit is input to the discrete cosine transformer Q1. At this time, the switch S1 switches the compression mode between the intra-frame compression mode and the inter-frame compression mode. In the case of the intra-frame compression mode, the original image data is directly input to the discrete cosine transformer Q1. In the case of the inter-frame compression mode, the difference value between the decoded image data read from the prediction memory M and the original image data is input to the discrete cosine transformer Q1.

The discrete cosine converter Q1 receives the input 8 ×
The original image data or difference values for eight pixels are converted into DCT coefficients in an 8 × 8 array by discrete cosine transform. At this time, from the 8 × 8 DCT coefficients finally obtained, the 5 × 5 low-frequency-side array elements are converted values from the original image data or the difference values and the remaining Are output as simple DCT coefficients in which all 0 (zero) are set in the high-frequency side array elements of. In addition, the second operation unit Q1-Q2
Outputs a DCT coefficient in which the original image data or the conversion value from the difference value is set in all of the 8 × 8 array elements.

Second operation unit Q1 of discrete cosine converter Q1
The output from -Q2 is input to the first quantizer SMP1. The first quantizer SMP1 quantizes the DCT coefficient to an 8-bit value (0 to 255). Then, the first quantizer S
The output from MP1 is input to the first entropy encoder C1, is subjected to variable-length encoding, and is output as first compressed image data.

[0039] The output from the second arithmetic unit Q1-Q2 of the discrete cosine transformer Q1 is also input to the inverse quantizer SMP1 ^-1. The inverse quantizer SMP1 ^-1 and the inverse discrete cosine transformer Q1 ^-1 are connected to the discrete cosine transformer Q1 and the first
Decoding is performed in a procedure reverse to the coding procedure by the quantizer SMP1, and the decoded image data is stored in the prediction memory M.
Individually, the decoded image data stored in the prediction memory M is
In the case of the above-mentioned inter-frame compression mode, it is used to obtain a difference value from original image data located before and after on the time axis.

On the other hand, the output from the first operation unit Q2 of the discrete cosine transformer Q1 is input to the second quantizer SMP2. At this time, only when the switch S3 is switched in conjunction with the switch S1 and the compression mode becomes the intra-frame compression mode, the simplified DCT count output from the first arithmetic unit Q2 is input to the second quantizer SMP2. .

The second quantizer SMP2 quantizes the simplified DCT coefficient into a 7-bit value (0 to 127). Then, the output from the second quantizer SMP2 is input to the second entropy encoder C2, subjected to variable-length encoding, and output as second compressed image data. It should be noted that among the configurations described above, the discrete cosine transformer Q1 to the first quantizer SMP1
The system that reaches the first entropy coder C1 via the above corresponds to the first encoding processing system according to the present invention. Further, a system from the first operation unit Q2 in the discrete cosine transformer Q1 to the second entropy encoder C2 via the second quantizer SMP2 corresponds to a second encoding processing system according to the present invention.

According to the image coding apparatus configured as described above, the discrete cosine transformer Q1 outputs both the DCT coefficient and the simplified DCT coefficient, and generates the first compressed image data based on the DCT coefficient. Since the second compressed image data is configured to be generated based on the simplified DCT coefficient, the second compressed image data is a data having a lower spatial resolution and a smaller data amount than the first compressed image data. Become.

Further, the first quantizer SMP1 stores the data as 8
Quantize to a bit value (0-255) and convert to a second quantizer SM
Since P2 is configured to quantize the data to a 7-bit value (0 to 127), the second compressed image data is a data having a lower amplitude resolution and a smaller data amount than the first compressed image data. Become.

Further, since the switch S3 is interlocked with the switch S1, the simple DCT count is inputted to the second quantizer SMP2 only when the compression mode is set to the intra-frame compression mode. The second compressed image data has a lower time resolution and a smaller data amount than the first compressed image data.

Therefore, according to this image encoding device, the first compressed image data having a low compression ratio and the compression ratio can be transmitted to two or more different transmission paths and output destinations without preparing two image encoding devices. , The second image data having a high value can be output in parallel. Specifically, for example, it is possible to transmit the first compressed image data to a digital broadcast network having a high transmission rate and simultaneously transmit the second compressed image data to a network network having a low transmission rate. Further, while the first compressed image data is recorded on the recording medium, the second compressed image data can be simultaneously transmitted to a network having a low transmission rate.

Further, in this image coding apparatus, a simplified DCT coefficient is generated by selectively extracting a part of the DCT coefficient in the discrete cosine transformer Q1.
A part of the configuration for generating the DCT coefficient and a part of the configuration for generating the simplified DCT coefficient are shared.
Compared to a configuration that independently has a configuration for generating a T coefficient and a configuration for generating a simplified DCT coefficient,
The apparatus configuration can be simplified and the apparatus configuration can be reduced in cost.

Although the embodiment of the present invention has been described above, the present invention is not limited to the above-described specific embodiment.
The present invention can be implemented in various other modes. For example, in the above-described embodiment, the second compressed image data reduces the spatial resolution by using the simple DCT coefficient, decreases the temporal resolution by using the switch S3, and
Although the amplitude resolution has been lowered by using a quantizer, some may be omitted if the desired compression ratio can be obtained without using all of them.

Specifically, for example, as shown in FIG.
The switch S3 may be omitted. In this case, the second compressed image data has a higher temporal resolution than that of the above-described embodiment, and can reproduce a moving image that moves more smoothly. Moreover, although the second compressed image data has a lower spatial resolution and lower amplitude resolution than the first compressed image data, the decoded image data for performing inter-frame compression uses the first compressed image data to obtain a difference value. Therefore, the quality of the inter-frame compressed image in the second compressed image data is less likely to be degraded than that in which the second compressed image data having a high compression rate is decoded and subjected to inter-frame compression. In addition,
A part of the configuration for generating the DCT coefficient and a part of the configuration for generating the simplified DCT coefficient are the same as in the above-described embodiment.

As shown in FIG. 4, instead of replacing information in the middle of the DCT operation processing with 0, a DCT coefficient of a two-dimensional matrix is obtained once, and arbitrary row components and column components are replaced with 0. A similar effect can be obtained by outputting a frequency component lower than the DCT coefficient of the first compressed image data via the dimensional limiter.

FIG. 5 shows the conversion process at that time. Alternatively, the two-dimensional limiter may be deleted, and the same DCT coefficient may be used for both the first compressed image data and the second compressed image data without using the simple DCT coefficient. In this case, the second compressed image data has a higher spatial resolution than that of the above embodiment. Further, since a configuration for generating a simple DCT coefficient is not required, the configuration of the apparatus is further simplified.

Further, in the above embodiment, an image compression apparatus capable of outputting two kinds of compressed image data having different compression rates in parallel has been exemplified. However, it is configured such that three or more kinds of compressed image data can be output in parallel. You may. Specifically, for example, in the above embodiment, the 5 × 5 low-frequency side array element is extracted from the 8 × 8 DCT coefficient to generate a simple DCT coefficient, but the 6 × 6 low-frequency If another simple DCT coefficient is generated by extracting the side array elements, the third compressed image data can be generated based on the other simple DCT coefficient. In addition, as described above, since the spatial resolution, the time resolution, and the amplitude resolution can be adjusted independently,
A plurality of different types of compressed image data can be output by adopting a plurality of types of encoding processing systems, such as one in which one is dropped, one in which two are dropped, and one in which all are dropped. Become like

[Brief description of the drawings]

FIG. 1 is a block diagram of an image encoding device.

FIG. 2 is an explanatory diagram showing conversion progress of DCT coefficients and simplified DCT coefficients.

FIG. 3 is a block diagram of an image encoding device in which a part of the configuration is changed.

FIG. 4 is a block diagram of an image encoding device in which another configuration is partially changed.

FIG. 5 is an explanatory diagram showing the progress of conversion of DCT coefficients and simplified DCT coefficients when another part of the configuration is changed.

[Explanation of symbols]

Q1: discrete cosine converter, Q2: first operation unit, Q1-Q2: second operation unit, SMP1: first
Quantizer, SMP2 · · · second quantizer, C1 · · · first entropy encoder, C2 · · · second entropy coder, SMP1 ^-1 · · · inverse quantizer, Q1 ^-1 ...
Inverse discrete cosine transformer, M ... prediction memory, F ...
Loop filter, S1 to S3 ... switch.

Continued on the front page F-term (reference)

Claims (5)

[Claims] 1. A plurality of compressed image data compressed at a plurality of compression ratios are generated by dividing original image data input from a single input unit into a plurality of encoding processing systems and encoding them. An image encoding apparatus for outputting each compressed image data in parallel from a plurality of output units, wherein at least a first encoding processing system encodes the original image data, and a data amount is smaller than the original image data. And generating at least a second compressed image data from the intermediate data partially processed by the first encoding processing system. And selectively generating second compressed image data having a data amount smaller than the first compressed image data based on the extracted partial data. apparatus. 2. The method according to claim 1, wherein the first encoding processing system converts the original image data in a two-dimensional array into DCT coefficients in a two-dimensional array by discrete cosine transform, and performs the first compression based on the DCT coefficients. Generating image data, wherein the second encoding processing system converts the DCT coefficients converted by the first encoding processing system into low-frequency-side arrays in a frequency array range lower than the frequency array range of the DCT coefficients; Elements are selectively extracted, and all array elements other than the low-frequency side array elements are set to 0 (zero) to simplify DCT.
The image encoding apparatus according to claim 1, wherein coefficients are generated, and the second compressed image data is generated based on the simplified DCT coefficients. 3. The first encoding processing system quantizes the DCT coefficient, and based on the quantized coefficient, generates the first DCT coefficient.
, And converts the quantized coefficients into decoded image data by inverse quantization and inverse discrete cosine transform, and performs inter-frame compression based on the decoded image data and the original image data. The method according to claim 2, wherein the second encoding processing system generates the simplified DCT coefficients from the DCT coefficients subjected to the inter-frame compression by the first encoding processing system. The image encoding device according to claim 1. 4. The first encoding processing system generates the first compressed image data that has been subjected to intra-frame compression and inter-frame compression, and the second encoding processing system generates the first compressed image data. Data corresponding to the frame subjected to only the intra-frame compression is selectively extracted from the intermediate data partially encoded by the encoding processing system, and the second data is extracted based on the extracted data.
2. The compressed image data of claim 1 is generated.
Alternatively, the image encoding device according to claim 2. 5. The first encoding processing system performs quantization by dividing intermediate data generated in the course of encoding by a first divisor, and the second encoding processing system The quantization is performed by dividing intermediate data generated during the encoding process by a second divisor that is larger than the first divisor. The image encoding device according to claim 1.