JP3009434B2 - Data compression device - Google Patents

Description

The present invention relates to a data compression apparatus in which compression processing units having different compression methods are connected in multiple stages, and a data compression apparatus capable of improving a compression ratio and suppressing a processing time. The compression processing unit in each compression stage includes a compression unit that compresses data held in the input data holding unit by a compression method provided in the stage, and a compression result holding unit that holds a compression result. Determining means for evaluating the compression ratio based on the data held in the input data holding means and the data held in the compression result holding means; and the data output to the next compression stage based on the evaluation result of the compression rate by the determining means. The output of either the input data holding means or the compression result holding means is selected, and the compression method selected for the compressed data output from each compression stage to the next stage is displayed. And selecting means for adding the data to be indicated.

[Industrial application field] The present invention relates to a data compression apparatus in which compression processing units having different compression methods are connected in multiple stages.

2. Description of the Related Art In recent years, with the spread of large-capacity memories and the improvement of computer capacity, large-capacity data such as image data has been increasingly handled. Along with this, a data compression device capable of realizing a high compression ratio so as to be able to store and transmit a huge amount of data in a storage medium having a small capacity as much as possible in terms of management of data movement and retention has been required. Was.

[Related Art] Various methods have been proposed as conventional compression methods for compressing data such as images and documents stored in files. For example, for images (graphics, video, etc.), the corresponding coding method, for example, run-length coding, modified Huffman (MH) coding, and modified read (MR) coding for scanned images such as facsimile pictures Huffman coding and the like for character code data such as inter-frame coding and intra-frame coding for television images, etc., each of which is a known method and is currently being studied. In addition, international organizations (CCITT, etc.) are considering it.

There are several conventional methods for obtaining compression of data having such a high compression ratio, examples of which are shown below.

Develop a single, high-performance compression scheme.

Simple compression or code modulation is performed as preprocessing to improve the compression ratio.

Apply various compression methods and select the most suitable one.

Various compression methods are applied uniformly to improve the compression ratio (in this case, the various compression methods are executed in multiple stages in the determined order anyway).

[Problems to be Solved by the Invention] Regarding the above-mentioned conventional technology, development of a method capable of realizing a high compression ratio cannot be realized immediately after a long period of time. In the method (1), target data that can achieve a high compression ratio is limited (for example, limited to images). In other methods, processing takes time. Further, in some cases, the data length after compression may exceed the original data length.

The problems (1) to (4) are always problems in performing the compression processing, and these problems are often contradictory to each other. There has not been an efficient compression method capable of storing files including text data such as images and documents as efficiently as possible and transmitting large-capacity data in a short time.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a data compression apparatus capable of improving the compression ratio of data of a file including an image and suppressing the processing time.

[Means for Solving the Problems] FIG. 1 is a block diagram showing the principle of the present invention.

In FIG. 1, reference numeral 10 denotes original data, 11 denotes a compression processing unit for performing a compression process according to a method (A, B,...) Determined corresponding to each stage, 12 denotes a compression means for performing data compression, 13 is input data holding means, 14 is compression result holding means, 15 is judgment means, 1
Reference numeral 6 denotes a selecting means for selecting output data to the next stage.

The present invention evaluates the compression ratio as a result of compressing data inputted by a certain compression method, and selects a compressed data if a certain compression ratio is present, and otherwise selects data before the compression is performed. In the next stage, the same processing is performed on the data to sequentially perform multi-stage compression.

[Operation] In FIG. 1, the original data 10 is stored in a file containing an image, data of the minimum processing unit is extracted from the file, and a compression processing unit is used in accordance with this method (first method A).
The data is supplied to an input data holding unit 13 in the compression unit 12. The result of compression by this method A is output to the compression result holding means 14 and held. When the compression processing is completed, the determination means 15 obtains a compression ratio (input data length / data length of the compression result) from the data length held in the input data holding means 13 and the data length held in the compression result holding means 14. Then, whether or not a predetermined compression ratio has been obtained (the compression ratio has been improved) is evaluated, the result of the determination is output to the selection means 16, and data to be supplied to the compression means 12 at the next stage is selected. In this case, when the compression ratio is improved, the compression result holding means is used.
14 is selected, and if the output is not improved, the output of the input data holding means 13 holding the data before compression is selected.
The data is supplied to the compression processing unit 11 of the next system B.

The compression means 12 of the compression processing unit 11 at the next stage performs the compression by the method B different from the method A of the previous stage by the same operation as described above, and the processing by each method is executed sequentially thereafter.

When the data of the compression result holding means 14 is selected in the compression processing by each of the above methods, data representing the compression method is added to the header (head portion) of the data of the compression result, and the compression processing is performed. Each time, the contents are added. As a result, the contents and order of the processing when restoring the original data are displayed.

Further, when the processing unit of the data to be processed by the compression means of each system is divided and processed by a part (minimum processing unit) of all data, the processing in each processing unit can be operated in parallel, and the processing speed is improved.

[Embodiment] FIG. 2 is a system configuration diagram of the embodiment, FIG. 3 is a configuration diagram of each processor module, FIG.
The figure shows the processing flow of the embodiment, and FIG. 6 shows the structure of packet data for compression processing.

In the case of this embodiment, the order of the compression schemes is flexibly made possible by a simple network using packets in the system configuration.

In FIG. 2, a network 20 for performing packet communication is shown.
And the local and
The local processor module 21 performs compression processing according to the method A, the local processor module 22 performs compression processing according to the method B, and the main processor module that controls the whole. The following local processor module
23 and 24 perform compression processing in different formats X and Z, respectively, and the main processor module 25 has a film 251 and
A file including a video is stored, and data for each minimum processing unit (for example, 2 Kbytes, 4 Kbytes or the like, which is determined by the memory capacity, etc.) is transmitted as packet data to the local processor module 21 as a destination. A, B... X, Z are known compression methods. The order of compression is arbitrarily determined. Further, each local processor module can execute high-speed processing by executing the processing of the divided unit in parallel in a pipeline manner.

FIG. 3 shows the configuration of each processor module such as a local processor and a main processor.

In FIG. 3, 30 is a receiver, 31 is a driver, 32
Is a CPU, 33 is a program memory storing a program for executing the respective functions (compression processing by a specific method, packet communication, compression ratio determination, etc.), and 34 is a memory for receiving and transmitting packet data. Data memory 35 is used for temporary storage of data in the compression process.
Memory.

The functions executed by the local processor modules (FIGS. 21 to 24) having the hardware configuration shown in FIG. 3 are shown in FIG. 4 as function explanatory diagrams.

In FIG. 4, packet data from the network is received by a receiving buffer (the receiving portion of the data memory 34 in FIG. 3), and its contents are processed in a processing block (FIG. 3).
The compression process is performed in the CPU 32 and the program memory 33), and the data of the compression result is output to the transmission buffer (the transmission portion of the data memory in FIG. 3). Compression judging unit (CPU 32, program memory 3 in FIG. 3)
In 3), the data length of the reception buffer and the data length of the processing result by the processing block are taken out, the compression ratio is determined, and when the compression ratio is improved, the contents of the transmission buffer (the compression result is already stored) are held as they are. If it does not improve, the contents of the reception buffer are overwritten on the transmission buffer (the previous contents are erased).

FIG. 5 shows a processing flow of the embodiment executed by the CPU 32 by the program memory 33 of the local processor module.

In this processing flow, as in the case described with reference to FIG. 4, data is first received in the reception buffer, and the data is compressed and output to the transmission buffer (50, 51 in FIG. 5).
Next, it is determined whether or not the compression ratio has improved (No. 52), and if not, the data in the reception buffer is transferred to the transmission buffer (No. 53). If it is improved, the compression method code is added and the name of the next processor is added (id. 54). Thereafter, transmission packet data is created by adding the processor name of the next stage to the data in the transmission buffer (55), and the packet data is transmitted to the network (56).

The structure of the packet data for the compression processing shown in FIG. 6 will be described.

The packet data shown in FIG. 6 is transmitted in order to make a request for compression to the next stage. FIG. 6A shows a destination code of the next stage (the local processor which performs the next stage processing).
The module number) and the following (b) represent the data size (the entire data length of the following (c) and (d)).

The following (c) shows data (compression method identification code, omitted if not compressed) representing the compression method in which compression was performed in this module. If compression has been performed by another module before input to this module, the compression method identification mode executed by this module is added after those compression method identification codes, and the history of the compression method is retained. You. Therefore, even if data is transferred to the next stage without applying a method having a low compression ratio, restoration can be performed.

The last (d) is the compression request data passed to the next stage. If compressed, data received from the previous stage (c) and (d) are data obtained as a result of compression. If not compressed, data received from the previous stage is received. Data of (c) and (d).

When the data compressed in this manner is restored, the data of (d) is referred to by referring to the applicable compression method identification code of (c) in the data of (c) and (d) obtained at the end. Is restored, the previous steps (c) and (d) are obtained by the restoration, and the restoration is further repeated to restore the original data.

[Effects of the Invention] According to the present invention, the processing speed can be increased for each minimum unit of compression by executing a plurality of compression schemes in a processing device capable of operating in parallel. Further, since the compression ratio is improved in the course of each compression process, a multi-stage compression processing result can achieve a high compression ratio.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the principle of the present invention, FIG. 2 is a system block diagram of the embodiment, FIG. 3 is a block diagram of each processor module, FIG. 4 is a diagram for explaining functions, and FIG. FIG. 6 is a flow chart showing the structure of packet data for compression processing. In FIG. 1, 10: original data 11: compression processing unit 12: compression means 13: input data holding means 14: compression result holding means 15: determination means 16: selection means

Claims (2)

(57) [Claims] In a data compression apparatus in which compression processing units having different compression schemes are connected in multiple stages, a compression processing unit in each compression stage performs compression on data held in input data holding means in each stage. Compression means for compressing according to the method, compression result holding means for holding the compression result, judgment means for evaluating the compression ratio based on the data held in the input data holding means and the data held in the compression result holding means, and compression of the judgment means. According to the evaluation result of the rate, either the output of the input data holding means or the output of the compression result holding means is selected as the data to be output to the next compression stage, and the compressed data output from each compression stage to the next stage is selected. Selecting means for adding data indicating the selected compression method. 2. The method according to claim 1, wherein the original data to be compressed is divided into fixed minimum units, and the results are sequentially transmitted to the next stage in a pipeline format that can be processed in parallel by each compression processing unit in each stage. A data compression device, characterized in that: