JPS63155889A - Picture data compressing device - Google Patents

Abstract

PURPOSE:To improve the clearness of a boundary line by providing an input picture memory, plural coding picture element registers, a predicting value calculator, plural data compressors, a comparator and a selector so as to decrease the section of fog near the picture boundary line. CONSTITUTION:An input picture stored in a memory 1 is extracted sequentially, stored in a register 3 as a coded picture element B, when a new value is stored in the register 3, the picture is transferred to a register 2. Then a data compressor 5 obtains a difference between outputs of the register 2 and the predicting value calculator 4, converts it into a transmission value, coded and outputted as a code A. Moreover, a data compressor 6 obtains a difference of outputs of the register 3 and the calculator 4, converts the result into a transmission value and coded and outputted as a code B. The value A is inputted to a comparator 7, compared with a fixed value from a signal line 9, the signal being the result of comparison is outputted to a selector 8, the selector 8 selects either the value A or the value B and gives an output to a signal line 10 as the code output. Thus, the section of fog near the boundary line is decreased to improve the clearness of the boundary line.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an image data compression system using a predictive coding method with information loss.

(Prior art) Typical data compression methods include a predictive encoding method that encodes the difference between a predicted value and a sample value, and a predictive encoding method that encodes the difference between a predicted value and a sample value, and a predictive encoding method that encodes the difference between a predicted value and a sample value. There is an orthogonal transform coding method that allocates fewer bits.

Comparing the two, the predictive encoding method has the feature that it can be implemented with a simpler device, and even in this predictive encoding method, if partial loss of information is allowed, significant data compression is possible.

In devices using predictive coding, the difference between the predicted value obtained from neighboring pixels and the pixel value of the hooded pixel is calculated, the difference value is converted into a transmission value, and the transmission value is encoded and data compressed. , transmission, storage, etc.

FIGS. 2 and 3 are examples of graphs for converting the difference value into a transmission value.

In these figures, the horizontal axis is the difference value, and the vertical axis is the transmission value.The fractions below the decimal point are rounded down during conversion.

Figure 2 shows an example of a conversion method in which the transmission value is fixed at a or -a when the absolute value of the difference is larger than a, and Figure 3 shows the value of the difference when the absolute value of the difference is larger than a. This is an example of a conversion method in which approximation is performed.

When converting a difference value into a transmission value in this way, values near 0 are usually converted accurately, and values with large absolute values with a low tζ occurrence rate are approximated or truncated at a certain value. be.

(Problems to be Solved by the Invention) Conventional predictive encoding methods have the disadvantage that images become blurred near boundaries where there is a large change in pixel values. The wider the change in pixel value, or the higher the data compression rate, the wider the range of blur.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image data compression device capable of shortening blurred sections near image boundaries and improving the clarity of the boundaries.

(Means for Solving the Problems) In order to achieve the above object, the image data compression device according to the present invention has two consecutive input images to be subjected to data compression.
means for layering and extracting pixels one by one as first encoded pixels and second encoded pixels; and means for decoding the code output that is the output of the data compression device, and a predicted value calculator that calculates a predicted value during encoding; a first data compressor that outputs a first code value that is the difference between the predicted value and the first coded pixel; a second code value that is the difference between the encoded pixel and the encoded pixel;
a data compressor; a comparator that determines whether the first code value is smaller than a predetermined value; and a selector for selecting the second code value.

(Example) Hereinafter, the present invention will be explained in more detail with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of an image data compression apparatus according to the present invention.

An input image to be data compressed is stored in the memory 1, and its pixel values are successively extracted in the horizontal or vertical direction and stored in the register 3 as encoded pixels B. Then, when the next new value is stored in register 3, the pixel stored until then is stored in register 2 as encoded pixel B.
Be moved to.

These code weighted amateurs and coded pixels B are coded using a predictive coding method, and the predicted values necessary at this time are calculated by the predicted value calculator 4 using the code outputs of pixels previously processed. Ru.

The simplest example is pre-prediction, in which the immediately preceding code output is decoded and used as a predicted value.

The data compressor 5 calculates the difference between the outputs of the register 2 and the predicted value calculator 4, converts the difference into a transmission value using the graph shown in FIG. Appear as a.

The data compressor 6 calculates the difference between the output of the register 3 and the predicted value calculator 4, converts the difference into a transmission value using a graph as shown in FIG. 3, and outputs it as a transmission value B.

The code value A is input to a comparator 7, where it is compared with a fixed value (for example, a in FIG. A signal indicating whether it is small or not is output.

The selector 8 selects either the code value A or the code value B according to the output of the comparator 7, and outputs it to the signal line 10 as a code output.

If the absolute value of code type A is smaller than a, code value A is selected; otherwise, code value B is selected.

The code output is not only transmitted or stored.

It is also input to the predicted value calculator 4.

(Effects of the Invention) As explained above, the present invention does not use this when there is a large difference between the predicted value and the actual pixel value, but uses the difference between the same predicted value and the next pixel value to code Since it is configured so that the boundary line is blurred, the blurred section near the boundary line can be shortened.

In particular, if the blurred section near the boundary line is only the pixel l-element on the boundary line that occurs during observation, if the device according to the present invention is used, pixels on the boundary line with intermediate values are ignored, so the difference code If the method shown in FIG. 3 is used and the approximation where the absolute value is large is suppressed, for example, so that the error is less than a, an image with clear boundary lines can be obtained as a restored image.

Furthermore, since the unique value of 0.1 pixel length is ignored in this case, it is also effective in removing noise on the image.

[Brief explanation of the drawing]

FIG. 1 shows an embodiment of the image data compression device according to the present invention, and FIGS. 2 and 3 show data compression m and converting the difference between the predicted value and the actual value of a pixel into a transmitted value. It is a graph. 1...Memory 2.3...Register 4...Predicted wage calculator 5.6...Data compressor 7...Comparator 8...Selector 9.10...Compensation line

Claims (1)

[Claims] means for sequentially extracting two consecutive pixels from an input image to be data compressed as a first encoded pixel and a second encoded pixel; a predicted value calculator that calculates a predicted value when encoding a coded pixel and a second coded pixel; and a first data compressor that outputs a first code value that is a difference between the predicted value and the first coded pixel. a second data compressor that outputs a second code value that is the difference between the predicted value and the second encoded pixel; and a comparator that determines whether the first code value is smaller than a predetermined value. and a selector that selects the first code value when the comparator determines that the code value is smaller than a predetermined value, and selects the second code value in other cases.