JPH06292026A - Picture data storage device - Google Patents

Abstract

PURPOSE:To compress picture data by a predetermined compression ratio by extracting a corresponding representative value in a storage means in response to each of stage data to be stored and providing an output as picture data. CONSTITUTION:Picture data 1 included in each picture processing unit divided by NXN pixels are divided into 1-K stages by a classification means 2 based on the value. Then the picture data of each pixel expressed by 1-K stage data are stored in a picture processing unit storage means 3. On the other hand, a representative value decision means 4 executes processing to decide a representative value from picture data groups of each stage and the representative value is stored in a stage representative value storage means 5. In the case of the output, a representative value output means 6 generates picture data 7 based on data stored in the means 3, 5 to provide an output. Through the constitution above, it is possible to compress the picture data 1 at a predetermined compression rate and the reproduced picture is made close to an original picture without a limit.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for compressing and storing halftone image information, and more particularly to a technique for always compressing at a constant compression rate.

[0002]

2. Description of the Related Art When displaying or printing a halftone image, it is necessary to save the density of the image in a storage medium in advance. For example, in order to be able to express halftones of 256 gradations in a black and white image, a storage device of 8 bits per pixel is required. In addition, if the image is in color, the storage requirements will be three or four times this amount.
Double (because it has to correspond to RGB or YMCB colors respectively). As described above, an extremely large amount of storage medium is required for an image data storage device that captures image data, stores the image data, and outputs the image data to a display device or a printing device.

Therefore, in order to efficiently store such a large amount of information, compression methods such as DPCM and JPEG have been proposed.

[0004] However, since these methods are the compression ratio according to the type of the record information to be place a compression would change, and the disadvantage that can not be predicted or in advance how much compression is applied, the cut-and-paste Even if an attempt is made to use possible graphic information, it is impossible because it is not possible to instantaneously determine at which position on the recording medium the information to be processed exists.

[0005] Therefore, to predict whether advance how much compression is applied, leaving the compression method of constant compression rate to what position in can be determined instantly on information storage unit to be processed for the idea Was done.

Above all, Mitsubishi Electric's “Fixed GBTC Algorithm
“M” divides the image data into blocks of 4 × 4 pixels, and divides the 16 pixels included in each block into 4 blocks.
It is classified into stages, and these values, the average value of the density data of one block, and the dynamic range are stored for compression storage. As a result, if 1 pixel has an information amount of 1 byte in the original image, 1 block requires 16 bytes, but 2 bits x 16 pixels + 1 byte + 1 byte requires 6 bytes.

[0007]

On the other hand, on the other hand, in the above method, it is sufficient that the distribution of the density data is close to the normal distribution on the principle of reproducing the original density data from the average value of the density data and the dynamic range. However, if the distribution is far from the normal distribution, the reproduced image has a drawback that it is significantly different from the original image.
Moreover, the type of image data that particularly needs to be compressed normally exhibits a distribution that is far from the normal distribution.

The present invention has been conceived in order to solve such a drawback, and an object of the present invention is to store image data representing halftone densities of pixels which are continuously sent and then display the image data. In an image data storage device for outputting to a printing device or a printing device, it is possible to compress the image data at a constant compression rate, and the reproduced image obtained by outputting the image data afterwards looks as close as possible to the original image. To do so.

[0009]

In order to solve such a problem, the image data storage device of the present invention uses image data contained in each image processing unit divided by N × N pixels from 1 to Based on the classification means for classifying into K stages and the classification result of said classification means, said N × N
Image processing unit storage means for storing individual image data of image processing units of pixels by replacing it with step data of 1 to K, and at each of the steps 1 to K, representative from the image data group included in each step A representative value determining means for determining a value and a representative value determined by the representative value determining means,
In accordance with the step representative value storage means for storing the step representative values in the order of 1 to K, and the step data stored in the image processing unit storage means, the corresponding representative value is stored from the step representative value storage means. Representative value output means for extracting and outputting this value as image data at the time of an output command.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a diagram showing the basic construction of the present invention.
In FIG. 1, first, input image data 1 is sent from the host computer or the like to the main image data storage device. The input image data 1 referred to here is, for example, data expressing the density gradation of a plurality of pixels forming a certain image, and is composed of 8 bits here. This allows
It is possible to express density gradation from 0 to 255.

Input image data 1 sent continuously
Is first divided into image processing units of 4 × 4 pixels, and processing is performed based on this unit.

Next, in this image processing unit, the input image data 1 is divided into four stages based on the classification means 2. By dividing into four stages, 8-bit information represented by 0 to 255 is converted into 2-bit information represented by 1 to 4. This time, the following method (1) was adopted as the classification method, but other methods such as (2) and (3) are also possible.

(1) Equal classification; 0 to 63, 64 to 1
27, 128 to 191, 192 to 255, etc.

(2) Emphasis classification on both ends; 0 to 42, 43
From 127 to 128, from 212 to 212, and from 213 to 255, a method of dividing the two on the inner side into a width of 2 when the width at both ends is 1.

(3) Central emphasis classification; for example, 0 to 8
3, 84 to 127, 128 to 171, 172 to 2
A method in which the width of the two ends is 2 when the width of the two inner parts is 1, such as 55.

The image data of each pixel converted into the 2-bit information represented by the step data of 1 to 4 is stored by the image processing unit storage means 3.

On the other hand, on the other hand, the representative value determining means 4 performs a process of determining a representative value from the image data group included in each stage at each stage from 1 to K. Various methods are conceivable for determining the representative value, but the most basic method is as follows.

(1) A method of taking an average value of all density data at the target stage.

(2) A method of taking the median value of all the density data of the target stage.

(3) A method of taking the most frequent value of all density data of the target stage.

In this embodiment, the method of taking the average value of (1) is adopted, and this value is written in the step representative value storage means 5 as a representative value.

As described above, it becomes possible to store data at a constant compression rate.

At the time of output, output image data 7 is created based on the data stored in the image processing unit storage means 3 and the step representative value storage means 5, and this data is output. In this process, a corresponding representative value is extracted from the stage representative value storage means 5 according to each stage data stored in the image processing unit storage means 3, and this value is output as output image data 7. For example, a pixel of an image processing unit has 172 as the input image data 1, which is 3
The stage representative value storage means 5 classifies the stage into three stages.
If it is determined that the representative value of the stage data of 165 is 165, 165 is output.

The above processing will be described more specifically with reference to FIG.

0 to 2 as image data for each pixel
The input image data 1 sent as an integer value within the range of 55 is divided into 4 × 4 image processing units 22 (a). In each image processing unit, the image data is 1,
It is classified into stage data 23 of 2, 3, and 4 (b). In each of the four stages, the average value is calculated from the four input image data 1 to which it belongs, and the stage representative value storage means 5 is created (c). At the same time, the image processing unit is filled with step data from 1 to 4, and the image processing unit storage unit 3
It is written to 1 (d). At the time of output, the representative value output means causes the step representative value storage means 5 and the image processing unit storage means 3 to operate.
To produce a 4 × 4 image processing unit for output, and transfer this to the output device.

Now, the image to be processed is 40 pixels x 4
Consider the storage capacity required to store information in the case of 0 pixels. The target image is divided into 100 image processing units, and each image processing unit 22 is divided into 16 image processing units.
Since there are pixels and 2 bits are needed for each pixel, 2 bits x 16 x 100 = 3200 bits = 400 bytes. On the other hand, the storage capacity required for the stage representative value storage means 5 is 8 bits (1 byte) for one of the representative values in the range of 0 to 255,
Since there are four stages, 4 bytes are required, so 4 bytes × 100 = 40 bytes. Therefore, according to the present invention, 400 bytes + 400 bytes = 800 bytes. However, according to the conventional method, 1 byte is required for 1 pixel, and since this is 1600 pixels, 1 byte × 160 = 1600 bytes. After all,
It can be seen that the present invention requires half the storage capacity to store image information as compared to conventional methods. This compression rate is 1
The number of pixels in one image is always constant, but if the image processing unit 22 is set to a value other than 4 × 4, or if the number of steps is set to 8 or 16 instead of 4, the compression rate will change. Of course.

As described above, according to the present invention, it is possible to store the image information sent with the density data of 0 to 255 with a smaller capacity as compared with the case where the image information is stored as it is without any processing. Become. In this case, the output image is not the original density data, but the representative value of each step is output, which is different from the original image.
However, the output image is faithfully reproduced to the original image invisible to the naked eye of the human being by finely dividing the steps into stages according to the degree of variation in the density data of the original image and accurately taking the representative value at each stage. It is possible.

[0028]

The image data storage device of the present invention is N × N.
Based on the classification result of the classification means 2 and the classification means 2 for classifying the input image data 1 included in each image processing unit 22 divided by pixels into 1 to K stages according to the value, the N × N Image processing unit storage means 3 for storing the individual input image data 1 of the image processing unit 22 of pixels by replacing it with step data of 1 to K, and an image included in each step in each of steps 1 to K. The representative value determining means 4 for determining the representative value from the data group, the step representative value storing means 5 for storing the representative values determined by the representative value determining means 5 in the step order from 1 to K, and the image. A representative value corresponding to each stage data stored in the processing unit storage means 3 is extracted from the corresponding stage representative value storage means 5, and this value is output as image data when an output command is issued. output By having a stage 6, it is possible to store the images at a fixed compression ratio.

[Brief description of drawings]

FIG. 1 is a diagram illustrating a basic configuration of the present invention.

FIG. 2 is a diagram for explaining the present invention with specific numerical values.

[Explanation of symbols]

 1 Input Image Data 2 Classification Means 3 Image Processing Unit Storage Means 4 Representative Value Determining Means 5 Stage Representative Value Storage Means 6 Representative Value Output Means 7 Output Image Data

Claims (2)

[Claims] 1. An image data storage device for storing image data representing halftone densities of pixels which are continuously sent and then outputting to a display device or a printing device. Each image processing divided by N × N pixels. A classifying unit that classifies the image data included in the unit into stages of 1 to K according to the value, and the individual image data of the N × N pixel image processing unit based on the classification result of the classifying unit, Image processing unit storage means for replacing and storing the step data of 1 to K; representative value determining means for determining a representative value from the image data group included in each step of the above 1 to K; Depending on the step representative value storage means for storing the representative values determined by the value determination means in the order of 1 to K, and the step data stored in the image processing unit storage means. And a representative value output means for extracting a corresponding representative value from the step representative value storage means and outputting this value as image data at the time of an output command. . 2. A compression rate is particularly halved by classifying the image data included in each image processing unit divided by 4 × 4 pixels into 1 to 4 stages according to the value in the classifying means. Claim 1 capable of
The image data storage device described.