KR0134322B1 - Noise reduction apparatus in ghost cancelling time - Google Patents

Abstract

The present invention relates to a noise reduction apparatus for removing ghosts that can prevent distortion of noise on the screen and improve processing speed when noise is mixed in a received ghost removal reference signal. . The present invention connects an intermediate value filter to an output terminal of a channel characterization processing unit which performs a process for obtaining a channel characteristic by obtaining a correlation between a received ghost elimination reference signal and a reference signal on the receiving side, thereby providing only intermediate values in the channel characterization data. The output will be filtered. The threshold comparison unit compares the filtered signal with a predetermined threshold to determine whether the input signal is a noise or a signal. The interference separation unit selectively outputs the channel characterization data and the height value according to the comparison result signal output from the threshold comparison unit to map data below the threshold determined as noise to zero.

Description

Noise reduction device when removing ghost

1 is a block diagram showing a first embodiment of the noise reduction apparatus in removing ghost of the present invention.

2 is a block diagram showing a second embodiment of the ghost noise reduction device of the present invention.

FIG. 3 is a block diagram showing a detailed configuration example of a subpixel generator and an intermediate value filter in FIG.

4 is a block diagram showing a detailed configuration example of a noise reduction device in removing ghost of the present invention.

5 is a block diagram showing a third embodiment of the noise reduction apparatus in removing ghost of the present invention.

5 is a block diagram showing a fourth embodiment of the noise reduction apparatus in removing ghost of the present invention.

6 is a block diagram showing a fifth embodiment of the noise reduction apparatus in removing ghost of the present invention.

7 is a graph showing original channel characterization data without noise.

8 is a graph showing channel characterization data using a ghost cancellation reference signal including noise.

FIG. 9 is a graph showing channel characterization data processed using a noise reduction device in ghost elimination according to the present invention shown in FIG.

* Explanation of symbols for main parts of the drawings

10: channel characterization processing unit 20: intermediate value filter

30, 60: threshold comparison unit 40: interference separation unit

50: subpixel generator 70: peak width detection and correction unit

The present invention relates to an apparatus for removing ghosts generated during transmission of an image signal. In particular, when noise is mixed into a received ghost removal reference signal, the noise is misinterpreted as a ghost and distortion appears on the screen. The present invention relates to a noise reduction apparatus for removing ghosts, which can prevent the occurrence of damage and delays of malfunction and processing speed.

As the current TV broadcast signal is a terrestrial broadcast signal, the signal is distorted due to the reflection of the transmission signal by the reflector, the distortion caused by the air and the non-ideal channel characteristics of the transmission medium, and the additional noise. As the TV broadcast signal passes through a multipath channel, a time delay difference and attenuation of a signal occur along a path, and as a result, a ghost phenomenon occurs in which multiple images are simultaneously overlapped on a main image of a TV screen. This ghost phenomenon has a greater impact on image quality than other additive noises. In particular, more serious problems may occur than in the case of information service media such as text multiplexing. That is, if a transmission code is misinterpreted as another code by a ghost signal, wrong letters or figures appear. The research on technology for removing such ghost phenomenon is steadily progressing, and among them, the Ghost Cancel Reference Signal is transmitted along with the transmission of the TV broadcast signal, and the ghost cancellation criterion is received at the receiving side. Research has been conducted on techniques for removing ghost components by analyzing the distortion of signals due to multipath channel characteristics by detecting signals.

When the image signal including the ghost elimination reference signal passing through the multipath is received at the receiver, the data of the vertical blanking interval lines containing the ghost elimination reference signal is read and classified in the sequence order of each field. . Subsequently, signal processing operations for waveform detection and channel characterization (CC) are performed to remove influences of the test signal, the synchronization signal, and the burst signal, and to obtain a ghost removal reference signal in which pure ghost is generated. In order to calculate the filter coefficient required for ghost removal using the resultant signal, that is, channel characterization (CC) data, an impulse of a channel capable of accurately detecting the number, magnitude, and delay of ghost generated through the channel characterization process You should get a response. The channel characterization process is classified into a differential operation method and a correlation operation method according to the type of the ghost removal reference signal, and the correlation operation method includes a linear correlation operation method and a cyclic correlation operation method.

However, if the noise is mixed during transmission during the waveform detection and channel characterization process at the receiving side ghost elimination device, the noise is incorrectly recognized as a ghost and processed, resulting in band-shaped pattern distortion on the screen. In addition, since the ghost cannot be detected quickly and accurately, the processing speed is delayed. Therefore, in order to secure the signal-to-noise ratio (S / N) for ghost elimination, a synchronous addition averaging (8 fields ^* n and n are integers) on a time axis or a finite impulsive response filter on a horizontal axis is conventionally performed. The low pass filtering is performed to limit the band, thereby affecting the noise. The synchronous addition averaging method is a method of averaging a digital signal sampling the same input signal for a predetermined number of times and applying it to a subsequent signal processing apparatus. As described above, averaging the input signals so that the sampling time points of the input signals coincide with each other is called a synchronous addition method. As the number of averaging times increases, the signal-to-noise ratio is improved. The low pass filtering method using the finite shock response filter compensates for the distortion by delaying the input digital signal at a predetermined time interval and continuously multiplying the delayed signal by a newly calculated coefficient. In this case, the output signal is represented by a linear combination of input data and its past values, and when the output signal is viewed on a time axis, the shock response is represented by a finite length.

However, according to the conventional method of averaging by synchronous addition, since the errors generated in the system itself cannot be eliminated, the averaged signal includes distorted information, thereby causing an error of inputting an incorrect signal into a subsequent signal processing apparatus. do. In addition, the method by finite impact response filtering of channel characterization data does not sufficiently suppress peak noise components of channel characterization data and suppresses original main peak signal (ghost elimination reference signal) and ghost peak signal. There was this. Since the noise not removed by this process is mistaken for ghost, a band-like pitonous distortion appears on the screen as it is. The distortion is severe when using a full tap filter that continuously updates all the coefficients of the filter, because the effect of noise in the channel characterized data is more sensitive in the flat constant region. Therefore, it is necessary to remove the noise component of the constant region. Since most of the channel-specific noise components are sub-peak components, it is necessary to remove the peak noise by nonlinear filtering rather than the linear filter finite impact response filtering.

SUMMARY OF THE INVENTION The present invention has been made to solve the conventional problems as described above, and an object of the present invention is to effectively suppress the peak noise component included in the channel characterization data by nonlinear filtering based on a medium filter. The present invention provides a noise reduction device for removing ghosts that prevents distortion on the screen and improves processing speed due to ghost misdetection.

In order to achieve the above object, the apparatus for reducing ghost noise according to the present invention implements a correlation between a received ghost cancellation reference signal and a reference signal on the receiving side, and performs a process for obtaining a multipath channel characteristic. Are respectively connected to the median filter and the interference separation unit. The median filter receives data from the channel characterization processor and filters the median from the input data. A threshold comparison unit is connected to an output terminal of the intermediate value filter, and the threshold comparison unit receives the filtered signal and compares it with a predetermined threshold to determine whether the input signal is a noise or a signal. The comparison result signal output from the threshold comparison unit is applied to the interference separation unit, and the interference separation unit selectively outputs data and zero values applied from the channel characterization processing unit according to the comparison result signal to zero data below the threshold value as a zero value. Will be mapped.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to FIGS. 1 through 9.

FIG. 1 is a block diagram showing a first embodiment of the noise reduction apparatus in removing ghost of the present invention. As shown, the channel characterization processing unit 10 for performing a channel characterization process for obtaining the multipath channel characteristics by comparing the received ghost cancellation reference signal with the reference signal in the receiver is performed by the intermediate value filter 20 and the interference separation unit ( Respectively. The median value filter 20 receives the channel characterization data from the channel characterization processor 10 and filters the median value from the input data and outputs the filtered median value. A threshold value comparator 30 is connected to an output terminal of the intermediate value filter 20, and the threshold comparator 30 receives the filtered signal and compares the filtered signal with a predetermined threshold value. The comparison result signal output from the threshold comparison unit 30 is applied to the interference separation unit 40, and the interference separation unit 40 according to the output signal of the threshold comparison unit 30 and zero channel characterization data ( By selectively outputting a zero value, data below the threshold is mapped to zero.

When the received ghost elimination reference signal is input to the apparatus of the present invention configured as described above, the channel characterization processor 10 obtains a correlation between the received ghost elimination reference signal and the reference signal of the receiver to obtain a characteristic of the channel. Will be processed to find. The processed channel characterization data is applied to the intermediate value filter 20 and the interference separation unit 40, respectively, and the intermediate value filter 20 passes the intermediate value only among the input data to limit the peak of the peak detail signal. Do The signal output from the intermediate value filter 20 is applied to the threshold comparison unit 30 and compared with a predetermined threshold value. When the input signal is less than or equal to the threshold value, the signal is judged as noise and outputs a low potential signal. If the input signal is larger than the threshold value, it is determined as a signal and outputs a high potential signal. The comparison result signal is applied to the interference separation unit 40, and the interference separation unit 40 outputs and maps a zero value when the comparison result signal has a low potential, that is, a part found to be noise by the comparison determination. When the result of the comparison is high potential, the original channel characterization data is passed to effectively suppress the noise component incorporated in the channel characterization data.

Here, the median filtering may use about 3 taps or 5 taps in consideration of hardware complexity. In the constant region, it is advantageous to increase the number of taps. However, since the reverse action may occur due to the loss of the peak value in the main peak signal or the ghost peak signal, the number of taps in the constant region and the number of taps in the peak portion may be adjusted differently.

7 is a graph showing original channel characterization data without noise. In the channel characterization data, the main signal by the ghost removal reference signal or the signal by ghost is shown in the peak signal form (A).

FIG. 8 is a graph showing channel characterization data using a ghost cancellation reference signal including noise. The channel characterization data, which is a secondary signal due to correlation, is a constant region signal (B) compared to a main peak signal (A) or a ghost piece signal. ) Are relatively more affected by noise, and malfunctions caused by noise are more sensitive to signals in the constant region.

FIG. 9 is a graph showing channel characterization data processed using a noise reduction apparatus of ghost elimination according to the present invention shown in FIG. As shown in the figure, when the noise peak is very large, residual noise C may remain even after the noise included in the channel characterization data is removed, and the low level value D around the main peak signal A is zero. By mapping, data may be lost.

2 is a block diagram showing a second embodiment of the noise reduction apparatus in removing ghost of the present invention. Here, the configuration and operation of each part of the noise reduction device are the same as those of the same reference numerals shown in FIG. However, by connecting a subpixel generator 50 between the channel characterization processor 10 and the median filter 20, an interpolation value between the data and the data applied from the channel characterization processor 10. By obtaining, the channel characterization data is made into more detailed data values and applied to the intermediate value filter 20. 2 is a configuration for correcting the loss of ghost having a small peak area.

FIG. 3 is a block diagram showing a detailed configuration example of the subpixel generator 500 and the median filter 20 in FIG. 2, and the median filtering is performed using channel characterization data as three-point subpixel data as an example. As shown, the subpixel generator 50 includes first and second sample delay elements 51 and 52 for delaying the channel characterization data applied from the channel characterization processor 10 at predetermined time intervals. A first adder 53 for adding the input channel characterization data and data output from the first sample delay element 51, and data output from the first and second sample delay elements 51 and 52; First and second dividers 55 for dividing the output signals of the first and second adders 53 and 54 by one-half times, respectively, to obtain subpixel data. And the intermediate value filter 20 includes the first sample delay element 51 and the first and second elements. Receiving the data values output from the period 55, 56, and outputs only the filtered intermediate value among them.

The subpixel generator configured as described above detects the current input data and the previously input data by delaying the input channel characterization data by one sample by using the first and second sample delay elements 51 and 52. In order to further refine the data, the first adder 53 adds the currently input data and previous data passed through the first sample delay element 51, and the first divider 55 adds the first adder 53. The median of the two signals input to the first adder 53 is obtained by dividing the output signal of the signal by 1/2. Similarly, the second adder 54 and the second divider 56 divide the second data by adding the previous data passing through the first sample delay element 51 and the second sample delay element 52 and doubling the second. The median of the two signals input to the adder 54 is obtained. The data output from the first sample delay element 51 and the first and second dividers 55 and 56 are all applied to the median filter 20, where only the median value is filtered and the threshold comparator. 30 is applied.

4 is a block diagram showing a third embodiment of the noise reduction apparatus in removing ghost of the present invention. Here, the configuration and operation of the respective parts of the noise reduction device have the same block draw as shown in FIG. In addition, the threshold comparison unit 60 is further connected between the channel characterization processing unit 10 and the intermediate value filter 20 to compare the input channel characterization data with a threshold set so as to reliably distinguish the presence or absence of ghosts. By setting the components below the threshold to zero, errors that may occur in subsequent processing can be eliminated in advance.

5 is a block diagram showing a fourth embodiment of the noise reduction apparatus in removing ghost of the present invention. Here, the configuration and operation of each part of the noise reduction device are the same as those of the same reference numerals shown in FIG. However, by further connecting the threshold comparison unit 60 between the channel characterization processing unit 10 and the subpixel generation unit 50, the input channel characterization data as in the third embodiment can be reliably distinguished from ghosts. By comparing the threshold value set to the threshold value and making the components below the threshold value zero, an error that may occur in subsequent processing can be eliminated in advance.

6 is a block diagram showing a fifth embodiment of the noise reduction apparatus in removing ghost of the present invention. Here, the configuration and operation of each part of the noise reduction device are the same as those of the same reference numerals shown in FIG. However, by connecting the peak width detection and correction unit 70 between the threshold comparison unit 30 and the interference separation unit 40, the width of the peak is detected and the threshold comparison unit 30 according to the detected value Correction of the output comparison result signal is performed. This is a means for eliminating the isolation point of the comparison result signal required for the noise separation unit 40 to map noise to a zero value. As shown in FIG. 9, when the noise peak is very large, the residual level C may remain even after removing the noise included in the channel characterization data, and all of the low level values D around the main peak signal Mapping to zero can adversely affect equalization. That is, medium value filtering with a large number of taps may be performed, or channel characterization data in which the comparison result signal becomes a high potential continuously may be counted and mapped to zero if the number is less than the set number. The main peak portion can be corrected by setting a window of a certain period that can include a low value (D) around the main peak and fixing the comparison result signal at high potential within the window tool. Although such a process is not shown by drawing, it can apply also to the structure of FIG. 2, FIG. 4, and FIG.

As described above, the present invention effectively removes peak noise components incorporated into the channel characterization data by nonlinear filtering based on the median filter, thereby reducing distortion on the screen by reducing ghost decision errors when searching for channel characteristics using the correlated peaks. This can be prevented from appearing, and the overall processing time can be reduced.

In addition, the intermediate value filtering, which is a nonlinear process, can be replaced by the minimum value filtering or the maximum value filtering.

Claims (11)

An apparatus for removing ghosts generated during transmission of a video signal, comprising: a channel characterization processor for performing a process for obtaining a multipath channel characteristic by obtaining a correlation between a received ghost removal reference signal and a reference signal on a receiving side; An intermediate value filter which receives data from the channel characterization processor and filters the median value in the input data and outputs the filtered value; A first threshold value comparison unit which receives the signal filtered by the intermediate value filter and compares the signal with a predetermined threshold value to determine whether the input signal is a noise or a signal; And a ghost separation unit for selectively outputting data and zero values applied from the channel characterization processor according to the comparison result signal output from the first threshold value comparison unit to map data below the threshold value to a zero value. Noise reduction device when removed. 2. The method of claim 1, wherein the interpolation value between the data applied from the channel characterization processor and the data is obtained between the channel characterization processor and the median filter, thereby making the channel characterization data into a more detailed data value and applying it to the median filter. Noise reduction device when removing the ghost, characterized in that configured by further connecting the sub-pixel generator. 3. The apparatus of claim 2, wherein the subpixel generator comprises: first and second sample delay elements for delaying channel characterization data applied from the channel characterization processor at predetermined time intervals; A first adder for adding the input channel characterization data and the data output from the first sample delay element; A second adder for adding data output from the first and second sample delay elements; And a first divider and a second divider for dividing the output signals of the first and second adders by one and a half times, respectively, to obtain subpixel data. The median filter of claim 3, wherein the median filter is a three-tap subpixel median filter that receives data values output from the first sample delay element and the first and second dividers, and filters only the median value. Noise reduction device when removing the ghost. 2. The method of claim 1, wherein the input channel characterization data is compared between a channel characterization processing unit and a subpixel generation unit with a threshold value set to reliably discriminate the presence or absence of ghosts, and the components below the threshold value are output to zero in advance. Noise reduction device when removing the ghost, characterized in that configured to further connect the second threshold value comparison. The method according to claim 2, wherein the input channel characterization data is compared with a threshold value set between the channel characterization processor and the subpixel generator to make it possible to reliably discriminate the presence or absence of ghosts, and the components below the threshold value are output to zero in advance. Noise reduction device when removing the ghost, characterized in that configured to further connect the second threshold value comparison. The comparison result of claim 1, wherein the width of the main peak signal included in the channel characterization data is detected between the first threshold comparison unit and the interference separation unit, and the comparison result is output by the first threshold comparison unit according to the detected value. Noise reduction device when removing the ghost, characterized in that configured to further connect the peak width detection and correction unit for correcting the signal output to the interference separation unit. 10. The method of claim 7, wherein the peak width detection and correction unit sets a window of a predetermined period that may include a low value around the main peak signal, and the comparison result signal output from the threshold comparison unit within the window section has a high potential. Noise reduction device when removing the ghost, characterized in that for fixing. The comparison result of claim 2, wherein the width of the main peak signal included in the channel characterization data is detected between the first threshold comparison unit and the interference separation unit, and the comparison result is output from the first threshold comparison unit according to the detected value. Noise reduction device when removing the ghost, characterized in that configured to further connect the peak width detection and correction unit for correcting the signal output to the interference separation unit. The comparison result of claim 5, wherein the width of the main peak signal included in the channel characterization data is detected between the first threshold comparison unit and the interference separation unit, and the comparison result is output from the first threshold comparison unit according to the detected value. Noise reduction device when removing the ghost, characterized in that configured to further connect the peak width detection and correction unit for correcting the signal output to the interference separation unit. The comparison result of claim 6, wherein the width of the main peak signal included in the channel characterization data is detected between the first threshold comparison unit and the interference separation unit, and the comparison result is output from the first threshold comparison unit according to the detected value. Noise reduction device when removing the ghost, characterized in that configured to further connect the peak width detection and correction unit for correcting the signal output to the interference separation unit.