JPS59135973A - Method and apparatus for transmitting still picture - Google Patents

Abstract

PURPOSE:To reduce the transmission time by transmitting roughly a picture at the first short time and transmitting detailed information of the picture as the time is elapsed in the transmission of a still picture having gradation. CONSTITUTION:A contrast picture signal is converted into a digital signal X, converted into a binary signal Y1 and stored in a memory 6, then inputted to a contrast estimation circuit 2 and converted into an estimate value X1. A contrast signal X subject to delay and compensation and its estimate value X1 are subtracted and the estimate error (X-X1) is stored in a memory 7 with binary coding. The binary signal Y1 stored in the memory 6 is coded under compression and modulated and transmitted to a transmission line 150. Then, the binary signal stored in the memory 7 is coded under compression and transmitted.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and apparatus for efficiently transmitting gray-scale still images by digital signal processing. Compression rate is 5 to 2 due to length encoding method etc.
0 efficient data compression encoding has been realized. On the other hand, differential coding (DPCM) and transform coding are known for images where the gradation changes continuously (ζ) such as photographs and television images, but the compression ratio is low at 2 to 8. , it takes a long time (1 to 2 minutes) to transmit still images over a telephone line.
, and efficient transmission had not been achieved.

SUMMARY OF THE INVENTION An object of the present invention is to provide a still image transmission method and apparatus that is simple and requires a short transmission time.

In the present invention, as a method for displaying shading, the outline of a shading image is first transmitted by applying the halftone method and dithering method used in newspapers and printing. Since the halftone image is expressed as a binary signal and can be significantly compressed, it is possible to transmit a rough image in a short period of time and inform the receiving side of the outline of the image content. Furthermore, since detailed information of the image is transmitted over time, it is possible to receive the image accurately over time.

According to the present invention, in transmitting a still image with gradation,
On the transmitting side, a binarization threshold f that periodically changes the image signal
The resulting binary signal Y1 is then encoded and transmitted to the first IC, and then the estimation error (X-XI) (XI
is the estimated value) and transmits it to the second, and on the receiving side, the first
decodes the code transmitted to obtain a binary signal Y1, converts it into an estimated value X1 for the image signal Xt, and displays it;
A still image transmission method is obtained which is characterized by decoding the transmitted code and displaying the calculated result with the estimated value X1. Means for binarizing using a periodically changing binarization threshold and the obtained binarized signal Y
1 to generate an estimated value X1 for the image signal xi to obtain an estimation error (XX+); means for encoding the binary signal Y1 and the estimation error (XX+); and a means for encoding the binary signal Y1 and the estimation error (XX+); a still image transmitting device comprising means for first transmitting an encoded signal Y1 and then transmitting an encoded estimation error; and a binary value that periodically changes a grayscale still image signal X. The signal Y1 is binarized from the conversion threshold l.
means for receiving a first encoded signal and decoding it into the original binary signal Y1; and means for converting the binary signal Y1 into an estimated value X1 for the still image signal Xi and displaying the converted signal. and means for decoding the second encoded signal in which the estimation error (XX+) is encoded next to the first encoded signal, adding the estimated value X, i, and displaying the result. A still image receiving device is obtained, which is characterized in that the first to fourth signals are formed from the above.

Next, the present invention will be explained in detail using the m-plane.

FIG. 1 is a block diagram showing an example of the configuration of an embodiment of a still image transmission device of the present invention. Hereinafter, the still image signal with gradations will be simply referred to as a grayscale image signal. In the figure,
The grayscale image signal is input from the terminal 100, and converted into a digital grayscale image signal X by the A/D converter 2 (H). Let it represent the shading of .

The grayscale image signal X is first binarized by a binarization circuit 1 whose threshold value changes periodically by f, and is converted into a binarized signal Y1. The binarized signal Y is stored in the first memo ++ 6, and is also input to the grayscale estimation circuit 2, and the grayscale signal xi
The estimated value X1f corresponding to this is transformed. The gray signal X whose delay has been compensated by the delay compensation circuit 3 and its estimated value X1 are sent to the subtracter 4.
The estimation error (X-XI) is extracted, and the estimation error (X-XI) is binarized by the second binarization circuit 1 and stored in the memory 71.

The binary signal stored in the memory 6 is sent to the data compression device 9.
It is compressed and encoded by the modulator 5, and then transmitted to the transmission path 150f.
This is sent. Next, the binary signal stored in the memory 7 is data compression encoded and transmitted. Here, the switch 8 is a switch for cutting off the signal for data compression.

In the control circuit 11, periodic binarization thresholds 101 and 101
', coordinate information 102, and switching signal 3 are generated. Although various methods such as a halftone method and a dither method can be applied to periodically change the binarization threshold value, a 4×4 systematic dither method will be explained here as an example.

FIG. 2 shows a 4.times.4 dither matrix, and the numbers indicate threshold values for binarizing a grayscale image signal having a level of 0 to 255 years. That is, 8 in the first column
．． 136.40.168 indicates the binarization threshold for the first scanning line, the first sample is 8, the second sample is 136,
The binarized closed value is 40 for the third sample and 168 for the fourth sample. From the fifth sample onward, these threshold values are repeated again. The binarization threshold for the second scanning line ζ is 20
0.72.232.104, and those of the third and fourth scan lines are 56.184.24.15.2 and 24
It is 8.120.216.88. The binarization threshold value is also periodically repeated in the sub-scanning direction.

When the signal Y1 obtained by periodically changing the threshold value of the grayscale image signal
It is well known that this shading can be reproduced. However, the reproduction accuracy of shading is poor.

Therefore, in the present invention, the binarized signal Y1 is first data compression encoded and transmitted, and the receiving side (sends rough image information, and then uses the binarized signal Y1 to transmit the grayscale image signal Xf). An estimated value X1 is generated, and the difference between X and Xl, that is, the estimation error, is data compression encoded and transmitted again, and a highly accurate grayscale image is finally transmitted.

FIG. 3 is a block diagram showing an example of the gray level estimation circuit 2. As shown in FIG. The binary signal Y1 is input to the tapped delay line 211, and the output is a binary signal yo- for 9 pixels consisting of 3 pixels in the main scanning direction and 3 lines in the sub-scanning direction, as shown in FIG.
ya is extracted. These 9-pin signals are stored in the ROM 22 (R,
ead 0nly Memory ) address line. ROM22f This is the center pixel of 3x3 pixels (
To memorize the corresponding estimated value X1 (from this,
Output (This 8-bit estimated value
It is preferable to calculate the statistics of the actual density values for the binarization pattern and adopt the density value with the highest probability of occurrence as the estimated value. The trouble with such a binary multi-value conversion method is as follows.
English magazine Proceedings of the
8ID+Vo 1' 22/3, 1981,
rDisplay of di on pp-185-189
Titled “thered imagesJ”, A, N.

Netravali and E, G, Bow2n
has been announced, so the details will be omitted.

In the present invention, it is also possible to separately provide a first encoder for data compression encoding the binary signal Y1 and a second encoder for data compression encoding the estimation error (X X+). In this case, the first encoder is a predictive a'l encoder designed for lased image signals (Ueno, Ono, Iwata, Inunishi, "Data Compression of Binary Dithered Images by Predictive Division Coding"). Child Communication Society Technical Research Report, IE78-47
(1978)) can be used. Also, the second
The encoder is a differential encoder for television signals (DPCM
encoder) can be used. Note that when using the second encoder, the multilevel signal (X-Xl) is directly stored in the memo IJ-7) without passing through the binarization circuit indicated by reference numeral 1'. For simplicity, an example is shown in which the first and second encoders are the same.

In Fig. 1, the signal line 1041 represents the estimation error CX-X+
) is supplied. Since this estimation error is a multilevel signal, this embodiment dither-encodes it again. Binarization circuit 11 changes periodically (X-X+)
Binarize. The estimation error has a positive or negative value, and
Since the corner distribution is concentrated, it is preferable to use, for example, the dithering shown in FIG. 5 as the binarization threshold.

In this way, the binarized signal Y2 is compressed and encoded using a data compression encoder having the same configuration as the first encoder, and is outputted to the transmission line 150 through the modulator 5. The switch 8 performs a switching operation to send out the second binary signal after the first 2-digitized signal has been sent out to the transmission line through the data compression encoder.

The compression rate of Data 4 varies depending on the type of image (this differs depending on the type of image, but roughly speaking it is 1. The first compression code can be compressed to 1/20 on average. , the amount of information without compression is 4 (10x 300x8
=960000 bits. It would take 100 seconds to transmit this over a telephone line using a 9600 bit/sec modem, but with the present invention, most of the image can be transmitted in the first 5 seconds.

Next, a method for receiving, decoding and reproducing the still image data compressed and encoded and transmitted will be described. In FIG. 1, a signal transmitted through a transmission line 150 is demodulated by a demodulator 15, and a first binary signal Y and l are decoded by a data expansion device 19g and stored in a memory 16. Ru. Binarized signal Y1 read from memory 16
is applied to the grayscale estimation circuit 12, and the estimated value Xlf of the grayscale image signal is converted into 0. Here, the grayscale estimation circuit 12 uses the same configuration as the one used here.

Since the binarized signal is stored, the ten views are -
This is processed by the second grayscale estimation circuit 121.
This is converted into a grayscale signal, which is added to the previously received and decoded grayscale estimated value X by an adder 141, output from the D/A terminal 2001, and displayed. Here, reference numeral 18 turns off the gate until the second encoded signal is received at the gate, sends zero to adder 14, and turns on the gate when the second encoded signal is received. do. Therefore, at the start of reception f, the grayscale estimated value X1 is displayed through the D/A converter.

Note that reference numeral 27 is a control circuit that outputs coordinate information 112, 112' and a gate switching signal 113 similarly to the transmitting side.

The second gradation estimating circuit 12' may have the same configuration as the gradation estimating circuit 2. The contents of the S1ROM are different because they are determined from the statistics of the estimation error number. If the estimation error in the gradation estimation circuit 121 is q, the output of the gradation estimation circuit 12' is (X-X1+q), so the output of the adder 14 is (X0%), and the reproduced image signal is the original gradation. The result is an image signal plus an estimation error.

In the above description, the second encoding is configured in one stage, but it can be configured in multiple stages.

In other words, the binary signal Y2 is converted into a gray signal by the further gray level estimation circuit, the estimation error with the Kugata signal is calculated, and this is dithered to obtain the binary signal ¥3. The error can be further reduced. FIG. 6 is a block diagram showing another embodiment of the still image transmitting apparatus of the present invention. This embodiment is characterized in that it consists of only one binarization circuit. The grayscale image signal is applied to a terminal 1001, converted into a digital signal by an A/D converter 20, and stored in a memory 261. This grayscale image signal It's the same. Binarization is performed by periodically changing the threshold value as explained in the dither method.

The binarized image signal is compressed and encoded by the data compression device 9 and sent out through the modulator 5 through the transmission path 150ζ, and is converted into an estimated value X1 for the image signal X by the grayscale estimation circuit 2I and stored in the memory 36. Stored. When the transmission of the first binary signal is completed in this way, encoding transmission of the estimation error (X-X+) is started. To do this, the estimated value X stored in the memory 36f
1 is read out, the gate 28 is turned on, and the subtracter 4c (CXX1) is taken out.0 This estimation error is binarized again in the binarization circuit 1, data compression encoded, and transmitted. Here, the second binarization threshold is the one shown in FIG. 5, and the gray level estimation circuit 2 is operated to create an estimated value for the signal (X−X, Add to X1, -・.

4 also stores the second estimated value X2 in the memory 36. X2 is an estimated value more accurate than Xl. By further repeating this operation, it is possible to transmit a signal with a higher degree of approximation. Note that reference numeral 29 is a control circuit that controls these operations.

FIG. 7 is a block diagram showing another embodiment of the still image receiving apparatus of the present invention. The present embodiment is characterized in that the receiving device is constituted by one grayscale estimation circuit. In the figure, compressed data demodulated by a demodulator 15 via a transmission line 150 is expanded by a data expansion device 19, and first, a first binary signal Y
I is decrypted.

The decoded binary signal is converted into a grayscale estimated value X+j by the grayscale estimation circuit 12F, and is stored in the memory 36 through the adder 14. Here, while receiving the first binarized signal, the adder 14 is turned off, so the adder 14 is turned off.
Add. The contents of the memory 36 are converted into an analog signal by a D/A converter and output to an output terminal 200 for display. Now, when the second binarized signal is received, the gate 31 is turned ON, and the estimated error (X-X+
) and the estimated value X1 already stored in the memory 36 is performed, and a more accurate grayscale image is stored in the memory and displayed. Reference numeral 39 indicates these control circuits.

FIG. 8 is a block diagram showing the configuration of a still image transmission device having both transmission and reception functions. It is comprised of a memory 26 for storing transmitted images, a memory 36 for storing received images, and a binarization circuit and a gray scale estimation circuit, respectively. Furthermore, this embodiment is configured so that it is possible to monitor the transmitted image at the time of transmission. That is, in the transmitting operation, the sent-out image signal Since the data is compressed by the data compression device 9 and supplied to the path 12 through the modulator/demodulator 25, the grayscale estimated value X1 for the binary signal being transmitted is stored in the received image memory 36. This is the same image that is sent to and displayed on the other station, and if the contents of the memory 36 are displayed through the D/A converter 30, your own station will see the same image as the other station. Become.

This local station monitoring function is especially useful for associations that use still image transmission equipment for long-distance conferences.

Now, when the transmission of the first binary signal is completed, the gate 41
is turned on and coded transmission of the estimation error (X X+') is started, and these operations are the same as described above. Further, in the receiving operation Cζ, the switch 42 selects the data expansion device 19 side and supplies the received binary signal to the gray level estimation circuit 12. Further, the reference number 49 is a control circuit for these.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the configuration of the present invention, Figs. 2 and 5 are diagrams showing an example of dither II socks, Fig. 3
The figure is a block diagram showing an example of a grayscale estimation circuit, FIG. 4 is a schematic diagram showing a pixel arrangement, FIG. 6 is a block diagram showing another embodiment of a still image transmitting device, and FIG. 7 is a block diagram showing another embodiment of a still image receiving device. FIG. 8 is a block diagram of a still image transmission device having a transmitting and receiving function. In the figure, reference numbers are 1.1'...binarization circuit, 2.12.12'...
... Grayscale estimation circuit, 3 ... Delay compensation circuit, 4 ... Subtractor, 5 ... Modulator, 6.
7.16, ]7...Memory, 8.42...
...Switch, 9...Data compression device, 15
0°°°°°・Transmission line, 15...Demodulator, 19
......Data decompression device, 14...Adder, 20...A/D converter, 30...D
/A converter, 18,28.31.41,...gate, 11,27.29.39.49...control circuit, 21...tapped delay line , 22-...
...ROM. 26.36,...me"" -,25...
...modulator/demodulator. Figure 2 Figure 4

Claims (1)

[Claims] 1. Transmission of a still image with gradation I, on the transmitting side,
The image signal X is binarized using a periodically changing binarization threshold, the obtained binarized signal Y1 is encoded and transmitted first, and then the image obtained using this binarized signal Y1 is The estimation error (X-X, ) (Xj is the estimated value) for the signal X is encoded and transmitted secondly, and on the receiving side, the first transmitted code is decoded to obtain a binary signal Y1, A still image transmission method characterized in that this is converted into an estimated value x1i for the image signal X and displayed, and then the second transmitted code is decoded and added to the estimated value X1 and displayed. . 2. Means for binarizing a still image signal X with gradation using a binarization threshold number that periodically changes the number, and generating an estimated value X for the image signal X using the obtained binarized signal Y1. means for obtaining an estimation error (X-X+);
1 and estimation error (X −X+ ); and means for first transmitting the encoded binary signal Y1 and then transmitting the encoded estimation error. Still image transmitting device 03 receives a first encoded signal obtained by encoding a signal Y1 that has been binarized by a binarization threshold value that periodically changes a still image signal means for decoding the signal Y1; and a means for decoding the binary signal Y1.
, a means for converting and displaying the estimated value X+l for the still image signal X;
-X+) when a second encoded signal is received, the device decodes the second encoded signal, adds it to the estimated value X1, and displays the result.