JPH02209098A - Picture transmission method - Google Patents

Abstract

PURPOSE:To attain the transmission of an excellent picture in a short period by applying thinning and interpolation, adding only a picture element with a large interpolation error and sending the result. CONSTITUTION:A so-called zigzag picture element only is read sequentially out of a memory 4 at a sender side, and the transmission by thinning is applied. The receiver side applies interpolation of picture elements to be sent. On the other hand, at the sender side, the interpolation is applied similarly at the receiver side. The calculated value is compared with the data of an actual picture element in the memory 4 and when the error is a permissible value or over, the address of the picture element is stored in a processor 5. When the transmission is finished, the address of the stored picture element is read and the address and the data of the picture element are sent in addition. To the sent data in addition, the receiver side detects the address and the data of the picture element of the address corresponding to the memory 17 is rewritten by the data sent in addition.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image transmission method suitable for use when transmitting an image signal using a transmission path corresponding to an audio band.

[Summary of the invention]

The present invention relates to an image transmission method, in which the transmission time is shortened by thinning and interpolation, and the transmission side performs calculations equivalent to interpolation, and only when the interpolation error is large, that pixel is additionally transmitted, thereby achieving good results in a short time. This allows image transmission to take place.

[Conventional technology]

The applicant of this application previously proposed a device that converts an image signal into an audio band and transmits it using a telephone line (Japanese Patent Application No. 62-1
61923).

[Problem to be solved by the invention]

However, in the above-mentioned apparatus, when converting an image signal into an audio band and transmitting it, it takes, for example, about 10 seconds to transmit a signal for one image (frame). Also, during this time, transmission of audio signals is interrupted.

Therefore, the user has to wait 10 seconds to view the image, and since the audio signal is cut off during this time, the user feels extremely uncomfortable during this waiting time.

This application has been filed in view of these points.

[Means to solve the problem]

In transmitting an image signal using a transmission path corresponding to the audio band, the present invention thins out the pixels of the image (memory (4)) to be transmitted on the transmitting side according to a predetermined rule, and then thins out the pixels on the receiving side during that time. The drawn pixels are interpolated and displayed, and the sending side performs the same interpolation calculation as the receiving side, compares the calculated value with the original pixel, and if the error is large, adds the original pixel. This is an image transmission method characterized in that the image is transmitted as follows.

(Operation) According to this, thinning and interpolation are performed, and only pixels with large interpolation errors are additionally transmitted, so that a good image can be transmitted in a short time.

〔Example〕

In FIG. 1, (1) is a video camera, and a video signal from this camera (1) is supplied to an AD conversion circuit (2), and the converted signal is sent to a memory (4) through a transmission processor (3). written to. The signal written in this memory (4) is read out through a transmission processor (3) at a low speed that can be converted into an audio band, and this read signal is supplied to a modulator (5) to modulate a predetermined audio signal. This modulated signal is sent to the switch (6) and the telephone line control unit (7).
) to the telephone line.

Also, the audio signal from the transmitter (8) is supplied to the amplifier (9), and the signal from this amplifier (9) is sent to the switch (6).
It is transmitted to the telephone line through the telephone line control section (7).

Furthermore, the audio signal from the telephone line is handled by the telephone line control section (7).
, are supplied to an amplifier (11) through a switch (10), and a signal from this amplifier (11) is supplied to a receiver (12).

At the same time, the signal supplied to the amplifier (11) is supplied to the detection circuit (13), and when it is detected that the signal from the telephone line is a modulated signal based on a video signal, the switch (10) is activated by the detection signal. is switched to the opposite position.

As a result, the signal from the telephone line is supplied to the demodulator (14), the demodulated signal is supplied to the AD conversion circuit (15), and this converted signal is written to the memory (17) through the reception processor (16). be included. This memory (17
) is repeatedly read out through the reception processor (16) at the speed of the video signal, the read signal is supplied to the DA conversion circuit (18), and the converted signal is sent to the monitor receiver and imager (19). ) and displayed.

In this device, the image written in the memory (4) is transmitted as follows.

First, from the transmitting side, only so-called staggered pixels, which are shown with 0 marks in FIG. 2, are sequentially read out from the memory (4) and transmitted by thinning.

Interpolation is performed on the transmitted pixels on the receiving side. That is, for example, 1 to 1 as shown in FIG.
Of the 9 pixels, the transmitted pixels ■, ■, ■,
By using the data d , d , d , d 1 in (2), it can be determined by the relational expression of the thinned out pixel mouth data. Note that this interpolation calculation is performed in accordance with the transmission of pixels, and this calculated data and the data of the transmitted pixels are written into the memory (17).

On the other hand, on the transmitting side, interpolation calculations similar to those on the receiving side are performed in parallel with the above-described transmission. This calculated value is then compared with the actual pixel data in the memory (4), and when this error value is greater than or equal to the allowable value, the address of that pixel is stored in the processor (5). Furthermore, when the transmission is completed, the stored pixel address is read out, and this address and pixel data are additionally transmitted.

The address of this additionally transmitted data is detected on the receiving side, and the data of the pixel at the corresponding address in the memory (17) is rewritten with the additionally transmitted data.

Image transmission is performed in this manner.

Therefore, regarding this transmission, the time required to transmit the first thinned out pixel is - when transmitting all pixels, and since it is thought that there are only a small number of pixels for which the error value is greater than the tolerance value, the overall The transmission time can be shortened to approximately - compared to the conventional method. Furthermore, by additionally transmitting data of pixels whose error values are greater than or equal to the allowable value, the resolution of the necessary portion is increased, and a good image can be obtained.

According to the above-described method, thinning and interpolation are performed, and only pixels with large interpolation errors are additionally transmitted, thereby making it possible to transmit a good image in a short time.

In more specific transmission, the image has a predetermined number of pixels (for example, 4X 4 = 16
), and in each divided block, for example, the pixel ■ at the upper left corner is transmitted the first time.

On the other hand, on the receiving side, the transmitted signal is memo IJ (17)
At the same time, the values at these positions are repeatedly read out and displayed within the blocks divided by the thick lines mentioned above.

As a result, this first transmission is performed in the conventional time of □, and the entire image can be displayed through this transmission, although the resolution is as low as □.

Furthermore, in FIG. 4, when the above-mentioned first transmission (A) is completed, each block becomes a small block consisting of fewer pixels (for example, 2X4=8) than the above-mentioned predetermined number of pixels, as shown in FIG. The divided small blocks are then transmitted a second time.

For example, if the block is horizontally divided into two small blocks as shown in the figure, the pixel l of the upper small block has already been transmitted, and in this second transmission, the pixel l of the lower small block has already been transmitted. Only the pixel ■ of the block needs to be transmitted.

As a result, this second transmission is also performed in the conventional time of □, and by this transmission, the entire image is displayed at the resolution of □.

Thereafter, the division is repeated sequentially as shown in C and D of the figure. In the fourth transmission, pixels (2) and (2) of each block are transmitted. When this fourth transmission is completed, the staggered pixels described above are transmitted, and the total transmission time at this time becomes approximately - of the conventional transmission time.

Therefore, according to such a transmission method, the user can see the entire image at an early stage, and the resolution of this image is gradually increased, thereby greatly reducing the discomfort of waiting time. It is. Note that in the above description, the size of divided blocks and the method of dividing small blocks are not limited to the above-mentioned example.

When such a transmission method using block division is adopted, it is possible to easily specify a portion where an error occurs when performing the above-mentioned additional transmission.

That is, in the above-mentioned apparatus, the portion where the interpolation error occurs is a high-resolution portion of the image, and such portions are generally considered to be concentrated at several locations in the image. Therefore, pixels with interpolation errors are determined in block units as described above, and an address is provided for each block.
The remaining pixels in the block are additionally transmitted together with this block address.

This eliminates the need to transmit addresses for each pixel, and reduces redundancy during transmission.

Furthermore, it is considered that blocks with such errors are concentrated in the image. Therefore, Fig. 5 shows an arbitrary image, and each section indicated by a solid line is a block resulting from the above division, and for example, as shown in the upper left corner, one section consists of 4 x 4 = 16 pixels. . For example, if it is detected that an error has occurred in a hatched range in the figure, a rectangular portion that includes this range and is indicated by a thick line is designated as the portion where the error has occurred. Furthermore, regarding this rectangular part, the block marked with a circle in the upper left corner is first identified by the serial number of the entire block, etc., and then the number of horizontal blocks (K) and vertical blocks that make up the rectangle with this block as the base point. The number (L) is determined. As a result, the serial number and the value of L are encoded and transmitted, so that the receiving side can determine the specified rectangular part from these codes and rewrite this part, reducing redundancy. can be further reduced.

By using the divided blocks in this manner, it is possible to efficiently specify the portion where an interpolation error has occurred.

In addition, in the above transmission method, if the number of pixels that are additionally transmitted is too large (the effect of reducing the transmission time will be weakened), then when the error value mentioned above exceeds the allowable value, the error value along with the address of that pixel will be It is also possible to store the data for the error in the processor (5) and to target only a predetermined number of pixels, starting from the one with the largest error, for additional transmission.

Further, when data for an error is obtained in this way, pixel data during additional transmission may be performed only for this error and added to the interpolated value on the receiving side. According to this, the number of bits of pixel data can be compressed.

Furthermore, the data of these pixels includes not only luminance information but also
It is also possible to include color signal information.

Furthermore, in the drawings described above, the transmitting side and the receiving side are drawn independently to simplify the explanation, but the configurations of the AD converting circuit (2), transmitting processor (3), and memory (4) are 15), a receiving processor (16) and a memory (1
7) and between the lines, etc., and these do not need to be operated at the same time, but are used in common in an actual device.

Furthermore, this transmission method allows transmission to be performed over a general analog telephone line, etc., and has a relatively high image transmission speed, so it is suitable for use in applications such as traffic monitoring systems at intersections.

〔Effect of the invention〕

According to this invention, thinning and interpolation are performed, and only pixels with large interpolation errors are additionally transmitted, thereby making it possible to transmit a good image in a short time.

[Brief explanation of the drawing]

FIG. 1 is a configuration diagram of an example of a device for carrying out the present invention;
2 to 5 are diagrams for explaining the present invention. (3) is a transmission processor, (4) is a memory, and (8) is a modulator. Moe Tsuru ■ Figure 5

Claims (1)

[Claims] When transmitting an image signal using a transmission path corresponding to the audio band, the pixels of the image to be transmitted are thinned out on the transmitting side according to a predetermined rule, and the thinned out pixels are thinned out on the receiving side. At the same time, the sending side performs the same interpolation calculation as the receiving side, compares the calculated value with the original pixel, and if the error is large, adds the original pixel and transmits it. An image transmission method characterized by: