JPS62122479A - Picture transmission system - Google Patents

Abstract

PURPOSE:To transmit efficiently a picture with high quality even in the transmission of a picture to a transmission system having a high error rate by sending data based on the difference between forecast data and original picture element data together with basic picture element data. CONSTITUTION:A sub-sampler 2 selects only prescribed picture element data in all inputted picture element data and gives an output. The basic picture element data outputted from the sub-sampler 2 is fed to an interpolation circuit 3 and a parallel/serial converter 8. The interpolation circuit 3 uses the basic picture element data to forecast the picture element data other than the basic picture element. The forecast data obtained by the interpolation circuit 3 is inputted to a subtractor 4 and data with the difference from the delayed signal by a delay circuit 6 is obtained. Data compressed by a nonlinear quantization section 5 is converted into a serial data string by a P-S converter, subjected to time division multiplex by a data selector 10 and the result is outputted from a terminal 13.

Description

DETAILED DESCRIPTION OF THE INVENTION <Field of Industrial Application> The present invention relates to an image transmission system, and more particularly to an image transmission system that performs high-efficiency encoding.

<Prior art> Conventionally, in systems that highly efficiently encode and transmit data obtained by sampling television signals, DPC is used.
Predicted differential code typified by M (differential pulse code modulation)
is widely used.

This is due to the fact that predictive differential encoding can sufficiently contribute to high efficiency with a relatively small amount of hardware.

<Problems to be Solved by the Invention> However, in the predictive differential encoding described above, if the number of quantization bits per pixel data is small, the l'F noise between two temporally consecutive screens will increase. The so-called edge business is likely to occur in areas where the image changes sharply due to variations, so it is not suitable particularly for high-efficiency encoding of high-quality images.

Furthermore, if the transmission system is a transmission system with a high error rate, such as a magnetic recording/reproducing system, this is not preferable because errors propagate due to the occurrence of data errors, causing extreme image quality deterioration.

In view of the problems such as Jll and 'A, the present invention provides an image transmission system that can easily transmit high-quality images to the next vehicle even when transmitting images to a transmission system with a high error rate. The purpose is to

Stage L for solving the above problems> For this purpose, the image transmission system of the present invention uses basic pixel data selected from all the pixel data constituting one screen to convert the remaining pixel data. The data is predicted based on the difference between the predicted data and the original pixel data and is transmitted together with the basic pixel data.

By configuring as described in L, if an error occurs in the transmission path, the error will spread only to the pixel where the error occurred or to pixels around it, making it possible to transmit high-quality images. became. Also, a quantization error of one pixel does not affect other pixels.

Edge business is also difficult to generate.

<Examples> Examples of the present invention will be described below with reference to the drawings.

1 to 3 are diagrams for explaining an image transmission system as an embodiment of the present invention, in which FIG. 1 is a schematic configuration of the transmitting side, FIG. 2 is a schematic configuration of the receiving side, and FIG. The figure is a diagram showing an example of the interpolation circuit in FIGS. 1 and 2.

In FIG. 1, a data string obtained by sampling, for example, a television signal at a predetermined sampling frequency and quantizing it to, for example, 7 bits is input to terminal 1 in FIG.

2 is a so-called sub-sampler, which selects and outputs only predetermined pixel data (hereinafter referred to as basic pixel data) from all the input pixel data.

FIG. 4 is a diagram showing an example of the arrangement of basic pixels. Here, the sampling frequency is an integer multiple of the water scanning frequency, and each pixel is arranged in a grid pattern.It is assumed that the pixels shown with diagonal lines in the figure are the basic pixels. The 7-bit basic pixel data is supplied to an interpolation circuit 3 and a parallel-to-serial (P-3) converter 8. It is assumed that this sub-sampler 2 is controlled by the output of a timing generation circuit 15 that generates various timing signals synchronized with a synchronization signal of input image data inputted from a terminal 14.

The interpolation circuit 3 is a circuit for predicting pixel data other than basic pixels using basic pixel data, and FIG. 3 is a diagram showing an example of the configuration of this interpolation circuit 3. The interpolation circuit shown in Fig. 3 constitutes a two-dimensional low-pass filter. For example, in Fig. 4, for the pixel indicated by XOI, the average value of basic pixels XOO and XO2, and for the pixel indicated by Xll, Basic image 2 XOO+ XO2* X 2 G r
The average value of X22 will be given as predicted data (interpolated value). In FIG. 3, 41 is a terminal to which the output of the sub-sampler 2 is supplied; 42 and 43 are IH scanning period (IH) delay circuits 44 and 45, respectively.

46.48 and 49.50 are coefficient multipliers, and 51° and 52 are l! ! : i period delay circuit, 47 and 53 are added respectively
rA, 54 is an output terminal.

The predicted data obtained by the interpolation circuit 3 is input to the subtracter 4, and the difference data between the input pixel data and the data delayed by the delay circuit 6, which delays the input pixel data by a period corresponding to the delay time of the interpolation circuit 3, is obtained. .

The occurrence distribution of the difference data (output from the subtractor 4) is shown in FIG. 5.As is clear from FIG. The number of quantization bits of data indicating the occurrence probability distribution can be reduced by performing nonlinear quantization.

The quantization characteristics of the nonlinear quantization circuit 5 are shown in FIG.
As is clear from the figure, the quantization characteristic is that when input data is near 0, it is finely quantized, and as the absolute value of the data becomes large, it is coarsely quantized. In the example shown in FIG. 6, when the data input from terminal l is 7 bits, the output is compressed to 3 bits.

The data compressed by the nonlinear quantizer 5 is transferred to a P-5 converter 9.
The basic pixel data obtained by sub-sampler 2 is also converted into a serial data string by P-5 conversion! i! 8, it is converted into a serial data string.

The serial data string output from the P-5 converter 8.9 is multiplexed by the data selector 10 by 11.1 times. In the output signal of the selector IO, the data rate is different between the basic pixel data (7 bits) and the compressed difference data (3 bits), which is not convenient for data transmission.
1, the data transmission rate is fixed. The synchronization addition circuit 12 is a circuit for inserting a synchronization signal so that the data positions of basic pixel data and compressed difference data can be identified on the receiving side, and is controlled by a timing generation circuit 15, such as a FIFO memory. A synchronizing signal is inserted into the output of 11 every period corresponding to 2 water scans,
The transmission data obtained thereby will be output from the terminal 13.

FIG. 7 is a diagram schematically showing the arrangement of transmission data. In the figure, 5 sync is a synchronization signal = xOO + Xo 2, etc. are basic pixel data, respectively, xo 1 + xo 3 + x10
+ XI I + XI 2 + xI 3, etc. indicate compressed difference data, respectively.

Next, the operation of the receiving side which receives this transmission data and restores the original pixel data will be explained.

In FIG. 2, received data supplied to terminal 'f-21 is separated into basic pixel data and compressed difference data by switch 21, and each is supplied to serial-to-parallel (s-p) converters 23 and 24. do. Since the rate of the parallel data output from the sp converters 23 and 24 is different from the original image data rate, the FIFO memories 25 and 26 return the parallel data to the original image data rate.

The FIFO memory 25 outputs 7-bit pixel data. The interpolation circuit 27 uses this basic pixel data to obtain predicted data (interpolated values) for pixels other than the basic pixels. Here, the interpolation circuit 27 uses the same circuit as the interpolation circuit 3 used on the transmission side.

On the other hand, 3-bit compressed difference data is output from the FIFO memory 26, and this 3-bit data is supplied to a nonlinear decoding circuit 28, which has characteristics opposite to the nonlinear quantization circuit 5 on the transmitting side, and is converted into 3-bit compressed difference data. This is the corresponding 7-bit representative value difference data.

The output data of the interpolation circuit 27 and the output data of the nonlinear decoding circuit 28 are added by the adder 29, and although they contain quantization errors due to nonlinear quantization, pixel data other than the basic pixels before encoding can be restored. It turns out.

The 7-bit basic pixel data output from the FIFσ memory 25 is output via the delay circuit 34 having a delay time corresponding to the delay time by the interpolation circuit 27, and is output from the adder 29 together with this basic pixel data. Data other than the basic pixels are output from a switch (selector) 30 in an order according to the position of each pixel, and restored image data is obtained from a terminal 33.

A synchronization signal included in the received data is separated by a synchronization separation circuit 31 and supplied to a timing generation circuit 32. Timing generation circuit 32 includes switches 22.30. It generates timing signals that determine the operation timing of each part of the S-P converters 23, 24'i.

In a transmission system configured as described above, 2. i! The propagation of error and the propagation of 9" CY- error occur only in a limited range. In this way, by preventing the propagation of errors, significant deterioration of the reproduced screen does not occur, so magnetic recording and reproduction It is possible to ensure reproduced image quality even when transmitting through a transmission system with a high error rate, such as a transmission system with a high error rate.In addition, because there is no propagation of quantization errors, edge business is less likely to occur, and the DPCM encoder Since recursive processing like the local decoder used is not required, when performing real-time processing on data that requires high-speed transmission, such as data obtained by sampling television signals, it is possible to use a so-called pipeline, etc. This is a simple method that can easily speed up the process.

In the above-mentioned embodiment, the basic pixels are one pixel every 2×2 pixels, but the interval between the basic pixels can be determined as appropriate based on the required image quality.

In addition, as the L method for obtaining predicted data for pixels other than basic pixels, predicted data is formed by two-dimensional average value interpolation, but it is also possible to form it by other methods. It is also possible to use the previous value prediction 3L, that is, the tree pixel data itself as prediction data, and to transmit the difference between the basic pixel data and data of other pixels.

<Effects of the Invention> As explained above, according to the present invention, it is possible to obtain an image transmission system that can efficiently transmit high-quality images even when transmitting images to a transmission system with a high error rate. It is.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a schematic configuration of a transmitting side of a transmission system according to an embodiment of the present invention. FIG. 2 is a diagram showing a schematic configuration of a receiving side corresponding to FIG. 1. Fig. 3 is a diagram showing an example of the configuration of the interpolation circuit in Figs. 1 and 2;
Figure 6 shows the frequency distribution of the output data of the subtracter in Figure 1, Figure 6 shows the quantization characteristics of the nonlinear quantization circuit in Figure 1, Figure 7 shows the frequency distribution of the output data of the subtracter in Figure 1, and Figure 7 shows the quantization characteristics of the nonlinear quantization circuit in Figure 1. FIG. 2 is a diagram schematically showing the arrangement of transmission data based on the configuration of FIG. In the figure, 3 is an interpolation circuit, 4 is a subtracter, 5 is a nonlinear quantization circuit, and xOOr XQ 2 and X20 * X22 are basic pixels.

Claims (1)

[Claims] An image transmission system that predicts remaining pixel data using basic pixel data selected from all pixel data constituting a screen, and transmits data based on the difference between the predicted data and original pixel data together with the basic pixel data.