JPH0435949B2 - - Google Patents

Description

[Detailed Description of the Invention] The present invention relates to a color image data encoding device,
In particular, it relates to image data in which, for example, color image data is encoded run by run.

When transmitting color image data of a multicolor image consisting of not only black and white but also a plurality of colors, the amount of information is much larger than when transmitting black and white image data. Therefore, color image data is compressed by encoding and then transmitted.

FIG. 1 is a block diagram showing an example of a conventional color image data encoding and decoding system. In the figure, color image data is input to an encoding device 1 in pixel sequential order. This color image data is given to the run detection circuit 2 and converted into run data. Here, a run is a series of pixels having the same color and arranged in the same scanning line direction. That is, the run detection circuit 2 converts the color image data into run length data (number of pixels in a run) and color data for that run. As a result, the color image data is considerably compressed. Furthermore, the output of run detection circuit 2 is applied to code conversion circuit 3 for data compression. The code conversion circuit 3 encodes the given run length data and color data,
Transmit. On the receiving side, the encoded data thus transmitted is encoded by the encoding circuit 4 and converted into the original color image data. Note that the decoding circuit 4 basically performs a conversion process opposite to that of the encoding circuit 1.

In the method shown in FIG. 1, color data must be encoded and transmitted for each run. Therefore,
The amount of code for color data increases, and transmission takes time. Furthermore, a memory with a large storage capacity is required to store code data.

Another method for reducing the amount of color-related information is to divide the screen into multiple blocks and color each block with a limited number of colors (for example, two colors). The power of expression is limited, and blocks must be taken into consideration at the stage of image creation, which is cumbersome.

Therefore, the main object of the present invention is to provide a color image data encoding device that obtains a small amount of color information without imposing any restrictions on the image.

To summarize, this invention encodes color image data for each pixel group of the same color unit, and includes:
The color data of a newly input pixel group is predicted based on the data related to the already input color image data, and if the prediction result is correct, only the data correlated to the pixel group position is encoded, and the prediction result is If there is an error, the color data and data correlated to the position are encoded.

The above objects and other objects and features of the invention will become more apparent from the following detailed description with reference to the drawings.

FIG. 2 is a block diagram showing one embodiment of the present invention. In the configuration, color image data is provided to a run length/color identification circuit 7 of an encoding circuit 5. This run length/color discrimination circuit 7 is connected to the run detection circuit 2 described above.
Similarly, color image data is converted into run length data and color data. The run length data and color data are provided to the prediction circuit 8 and also to the memory 9. The prediction circuit 8 predicts the color of a newly inputted run based on the data stored in the memory 9. Further, the run length data and color data are provided to a code conversion circuit 10. This code conversion circuit 1
0 encodes run length data and color data similarly to the code conversion circuit 3 described above. Here, the prediction circuit 8 controls the operation of the code conversion circuit 10 depending on whether the predicted result matches the actually input color data, that is, whether the predicted result is correct or incorrect. As will be described later, when the prediction result is correct, the code conversion circuit 10 encodes and transmits only the run length data. Furthermore, if the prediction result is incorrect, both run length data and color data are encoded and transmitted. The code transmitted as described above is given to the code identification circuit 11 of the decoding circuit 6. This code identification circuit 11 identifies whether the given code is a run length code or a color code, and outputs the run length code and color code separately. The run length code is given to a run length code inverse conversion circuit 12 and is inversely converted to the original run length data. Further, the color code is given to a color code inverse conversion circuit 13 and is inversely converted to the original color data. The outputs of the run length code inverse conversion circuit 12 and the color code inverse conversion circuit 13 are given to a prediction circuit 14. This prediction circuit 14 predicts the color of the run in the same manner as the prediction circuit 8 described above when only run length data is given. Note that the memory 15 already stores decoded run length data and color data, and the prediction circuit 1
4 makes a prediction based on this stored data. Then, the prediction circuit 14 outputs the color data predicted at this time and the given run length data as the data of the run at that time. Furthermore, when the prediction circuit 14 is given both run length data and color data, it outputs both of the given data as is as run data.

In operation, we now specify the color of run X to be encoded.
(X), and the color of the run predicted by the prediction circuit 8 is (X'). Also, if the run length of run X is (LX), then the code CDX of run X to be transmitted is
is expressed by the following equation.

CDXF(LX)(X)=(X') When G(X) and F(LX)(X)≠(X)(X') Where, F(LX) represents the run length (LX) G(X) is a code representing color (X).

Figures 3a to 3e are diagrams showing various run arrangements. In this manner, the run length data and color data of each run are stored in the memory 9. The prediction operation of the prediction circuit 8 will be explained below with reference to FIG.

Now, assuming that color A of run A is to be predicted, data of runs near run A are used for prediction. This uses the color correlation of runs in the image. Here, since it is clear that run B immediately before run A has a different color from run A, the data of run B is excluded from the prediction data.

First, as shown in FIG. 3a, the scanning line immediately before run A has a color different from run B, and run A
When there is only one run in contact with run C (runs with the same color as run B are excluded from the prediction data for the reason mentioned above. Below, Figures 3 b to e
The same goes for . ), color data C of run C is run A
predicted as color data.

Next, as shown in Figure 3b, if there are multiple runs such as runs C, D, and E that are in contact with run A and have a different color from run B, then The color data of the run with the largest number of pixels is predicted as the color data of run A. Note that instead of predicting the color of any run based on the number of pixels that are in contact with the run A, prediction may be made simply based on the position of the run that is in contact with the run of interest. For example, the color of the leftmost run C may be unconditionally the predicted color of run A, the color of the middle run D may be unconditionally the predicted color of run A, or the rightmost run The color of E may be unconditionally set as the predicted color of run A.

Next, as shown in FIG. 3c, if there is no run that is in contact with run A and has a different color from run B, then a run that has a color different from run B within a certain distance from run A in the scanning line direction. It is determined whether or not there is. If a run C is found within a certain distance, the color data C of this run C is predicted as the color data of run A. If there are multiple runs C and D having a different color from run B within a certain distance from run A in the scanning line direction, as shown in FIG. is predicted as run A color data. Note that instead of predicting the color of any run based on the distance from the run of interest, prediction may be made simply based on the position of a run within a certain distance in the scanning line direction from the run of interest. good. For example, run C on the left
You can unconditionally set the color as the predicted color of run A, or
Run D on the right side may be unconditionally predicted as run A.

Next, as shown in Figure 3e, there must be no run that is in contact with run A and has a color different from run B, and furthermore, there must be a run that has a color different from run B within a certain distance from run A in the scanning line direction. For example, color data C of run B immediately preceding run A and run C immediately preceding run A is predicted as the color data of run A.

Note that if the scanning line where run A is located is the topmost scanning line of the image, there may be no run used for prediction data in pixels near run A. In such a case, for example, the color of the background of the image is initially set in the memory 9, and this initially set color is used as prediction data.

4A and 4B are flowcharts for explaining the operation of the prediction circuit 8. FIG. The operation of this embodiment will be described below with reference to FIGS. 2 to 4B.

First, the run length/color discrimination circuit 7 converts color image data into color data and run length data. The color data and run length data are provided to the memory 9 and stored therein. The prediction circuit 8 also performs step 1 (abbreviated as S in the figure) of the flowchart shown in FIG. 4A.
According to operations 1 to 14, the color data of the run of interest A is predicted. Since this operation will be easily understood from the explanation of FIG. 3, further explanation will be omitted here.

Next, the prediction circuit 8 performs step 15 shown in FIG. 4B.
It is determined whether the run A is within a certain distance in the scanning line direction from the run A which has a different color from the run B. If they are within a certain distance, it is determined in step 16 whether or not a run having a different color from run B is in contact with run A. If it is determined that they are in contact, as shown in FIGS. 3a and 3b, the color of the run with the largest number of adjacent pixels is determined as the predicted color of run A in step 17. Then, in step 18, it is determined whether the actual color of run A is the same as the predicted color. If the color of run A is the same as the prediction, that is, if the prediction is correct, in step 19, the code conversion circuit 10 is instructed to only convert the code of the run length of run A. Accordingly, the code conversion circuit 10 transmits only the run length code to the decoding circuit 6. On the other hand, if the color is different from the predicted color of run A, that is, if the prediction is wrong, proceed to step 20.
At this point, the code conversion circuit 10 is instructed to convert the color and run length of run A. Accordingly, code conversion circuit 10 codes both the run length and color of run A and transmits them to decoding circuit 6. At that time, the code conversion circuit 10 transmits an identification code before the color code so that the color code transmitted when the prediction is incorrect is identified as a run length code. Note that instead of transmitting this identification code, the color signal itself may be a code that can be distinguished from the run length code.

On the other hand, if it is determined in the above-mentioned step 16 that a run having a different color from run B is not in contact with run A, that is, in the case shown in FIGS. 3c and 3d, the operation of step 21 is performed. That is, the color of the run with the shortest distance held in step 12 of FIG. 4A is taken as the prediction for run A. Below, the operations of steps 18 to 20 are performed. Furthermore, if it is determined in step 15 that there is no run with a different color from run B within a certain distance from run A, that is, the third
In the case as shown in FIG. The color of the run read out in step 23 is then set as the predicted color of run A. Subsequently, the operations of steps 18 to 20 described above are performed.

In the decoding circuit 6, the prediction circuit 14 performs prediction as explained in FIG. 3 only when color data is not provided from the color code inverse conversion circuit 13. Therefore, the prediction circuit 14 outputs data that is exactly the same as the color data and run length data output from the run length/color discrimination circuit 7.

As described above, when the color predicted by the prediction circuit 8 matches the color of the actual data, the color data of that run is not encoded or transmitted, so the amount of code to be transmitted can be reduced.

FIG. 5 is a block diagram showing another embodiment of the invention. In the configuration, the same parts as in FIG. 2 are given the same reference numerals, and the explanation thereof will be omitted. In this embodiment, memory 90 of encoder circuit 100 stores color image data as is. Then, the prediction circuit 80 converts the color image data read from the memory 90 into run length data and color data, which are used as prediction data. Furthermore, in the decoding circuit 600, the memory 150 stores color image data for each pixel, similar to the memory 90. Furthermore, the prediction circuit 140
whose output is pixel-by-pixel color image data,
Furthermore, the format of image data referred to in prediction is pixel by pixel.

In the embodiments described above, run data included in the scan line immediately before the run to be predicted is mainly used as reference prediction data, but run data included in multiple scan lines above the run to be predicted is used as the reference prediction data. data may be used as prediction data. For example, when using run data included in the upper two scan lines of target run A as prediction data, as shown in Figure 6, the run color is (E) = (C) ≠ (B) and (F )=(D)
When ≠(B), it is predicted that runs of the same color are connected in a downward-to-right direction to form an image,
It can be predicted that the color of run A is equal to the color of run C. In this way, by using runs of a plurality of scanning lines as prediction data, the configuration state of the image can also be taken into consideration, and the prediction matching rate can be further improved. Therefore, it is possible to further reduce the amount of code. In the case of the embodiment shown in FIG. 2, the color of run D, which has the largest number of pixels in contact with run A among the runs that are in contact with run A, is predicted as the color of run A.

Furthermore, in the embodiment described above, color data and run length data are used as data representing a run, but run break position (address) data may also be used instead of the run length data. good. That is, as long as the data correlates with the position of the run, distance calculations and pixel count calculations can be performed.

As described above, according to the present invention, only erroneously predicted color data is encoded, so that the code amount of color image data can be significantly reduced.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a conventional color image data encoding and decoding system. FIG. 2 is a block diagram showing one embodiment of the present invention. Third
The figure shows various arrangements of runs. 4A and 4B are flowcharts for explaining the operation of the prediction circuit 8. FIG. FIG. 5 is a block diagram showing another embodiment of the invention. FIG. 6 is an illustrative diagram for explaining an example of referring to runs of multiple scanning lines as prediction data. In the figure, 7 is a run length/color discrimination circuit, 8, 1
4, 80 and 140 are prediction circuits; 9, 15, 9
0 and 150 are memories, and 10 is a code conversion circuit.

Claims (1)

[Scope of Claims] 1. A color image data encoding device that encodes color image data input in pixel order, which divides the color image data into pixel groups of the same color, and determines the position of each pixel group. Data converting means for converting into color data indicating data correlated to and the color thereof; Storage means for storing at least part of the data related to the color image data that has already been input; Each pixel group output from the data converting means. prediction means for predicting color data based on data stored in the storage means; data encoding means for encoding color data and data correlated to the position output from the data conversion means; and the prediction. The control means controls the encoding operation of the data encoding means in accordance with the prediction result of the prediction means, and the control means combines the color data predicted by the prediction means and the color data output from the data conversion means. means for comparing and determining whether the two match/mismatch; and when the determining means indicates that the prediction result of the predicting means is correct, the data encoding means encodes only data correlated to the position. and, on the other hand, when the determining means indicates that the prediction result of the predicting means is incorrect, controlling the data encoding means to encode the color data and data correlated with the position. A color image data encoding device comprising means for encoding. 2. The color image data encoding device according to claim 1, wherein the data correlated with the position is run length data of each pixel group. 3. The color image data encoding device according to claim 1, wherein the data correlated with the position is address data indicating a boundary position of the pixel group.