KR960013227B1 - Method and apparatus for ghost canceller filter coefficient weighted average processing - Google Patents

Abstract

No content.

Description

Filter Coefficient Weighted Average Processing Method of Ghost Elimination Device and Its Circuit

1 is a block diagram showing the basic configuration of a general ghost removal device.

FIG. 2 is a block diagram specifically showing the configuration of the filter circuit shown in FIG.

FIG. 3 is a block diagram showing the configuration of the filter coefficient calculating circuit used in the ghost elimination device shown in FIG.

Figure 4 is a block diagram schematically showing a filter coefficient weighted average processing circuit of the ghost removing device according to the present invention.

5 is a block diagram showing a specific configuration of the weighted average processing unit shown in FIG.

6 is an operation flowchart of the filter coefficient weighted average processing circuit shown in FIG.

* Explanation of symbols for main parts of the drawings

10: A / D conversion circuit 12: reference signal extraction circuit

14: reference signal generation circuit 16: filter coefficient calculation circuit

18: filter circuit 20: D / A conversion circuit

22: delay unit 24: FIR filter

26: IIR filter 300: field operation circuit

305: correlation unit 310: reference point setting unit

315: IIR window wing part 320: FIR window wing part

325,330,360: FFT unit 340: first operation unit

345: second operation unit 350: third operation unit

355,370 IFFT part 365 Clamp part

410: FIR filter coefficient weighted average processing unit 420: IIR filter coefficient weighted processing unit

500: control unit 510: first working memory

520: weight setting unit 530: second working memory

540: weighted average operation unit

The present invention relates to a filter coefficient weighted average processing method and a circuit thereof of a ghosting device, and more particularly, to a filter coefficient calculated by a filter coefficient calculating circuit of a ghost removing device. The present invention relates to a filter coefficient weighted average processing method and a circuit of a ghost elimination device, which prevents a sudden change of a screen by optimally adjusting a change of a filter coefficient in consideration of a weighted average calculated threshold coefficient of.

In general, when a television signal is transmitted from a broadcasting station to a receiver through an air or air channel, it is reflected by a building or a terrain such as a mountain, or when the impedance matching of the line is improperly reflected, the original broadcast signal is received by the receiver. Apart from this, a so-called ghost phenomenon occurs in which the reflected signal appears on the screen.

Such ghosts are classified as pre-ghosts appearing before the original signal and post-ghosts appearing later than the original signal, and also as long ghosts appearing with a certain amount of time in the near ghosts and the original signal appearing nearby. can do.

Therefore, a ghost elimination device is employed to remove such ghosts. According to the principle of the ghost elimination device, a transmitter of a broadcasting station first transmits a ghost elimination reference signal (hereinafter referred to as a reference signal) at some time of a television signal. When the receiver has the same reference signal as that transmitted by the broadcasting station, the receiver determines the characteristics of the transmission path by comparing with the reference signal transmitted from the transmitter, and then configures a filter for compensating for the distortion of the signal. Passing through the filter compensates for distortion in the transmission path.

FIG. 1 is a block diagram showing a basic configuration of a general ghost elimination device. The ghost elimination device shown in the drawing includes an A / D conversion circuit 10 for converting an analog image signal input into a digital signal, and the digitally converted image. In the reference signal extraction circuit 12 for extracting the reference signal inserted at the time of transmission from the signal, the reference signal generation circuit 14 for generating the same reference signal as the reference signal inserted in the transmitter, and the reference signal generation circuit 14 A filter coefficient calculation circuit 16 for calculating a filter coefficient for compensating for distortion of a signal generated in a transmission path by comparing the generated reference signal with the reference signal extracted by the reference signal extraction circuit 12, and the A / D conversion A filter circuit 18 for receiving a digitally converted video signal from the circuit 10 and compensating for distortion occurring in a transmission path based on a filter coefficient calculated by the filter coefficient calculating circuit 16; It consists of a corrupt signal to the D / A converter circuit 20 for conversion into analog video signal.

FIG. 2 is a block diagram specifically showing the configuration of the filter circuit 18 shown in FIG. 1. The filter circuit 18 is a FIR filter (Finite-duration Impulse-) for removing a near ghost including all ghosts. A response filter 24, a delay unit 22 for delaying the original signal to remove all ghosts, and an IIR filter for removing long ghosts.

The filter coefficient calculating circuit 16 receives a received reference signal extracted by the reference signal extraction circuit 12 and a reference signal generated by the reference signal generating circuit 14 to remove a near ghost including all ghosts. The filter coefficients for the FIR filter 24 and the filter coefficients for the IIR filter 26 for removing the long ghost are calculated and output to the filters 24 and 26.

As such, the circuit structure for calculating the filter coefficient and its calculation method are already known and have been applied to practical use. The calculation time of the filter coefficient in the known technique is long, and the period of updating the filter coefficient is also long. Therefore, when the viewer frequently changes the broadcasting channel or changes the ghost due to some other reason, there is a problem that the removal operation of the ghost appears on the screen to the extent that the viewer feels.

In order to solve this problem, it is possible to reduce the calculation amount by correlating only a section of the input signal to obtain a peak value, while reducing the ghost elimination time by calculating the filter coefficient at once using the FFT (Fast Fourier Transform). A filter coefficient calculation circuit of a ghost elimination device has been proposed, and FIG. 3 is a block diagram showing an example of the configuration of the filter coefficient calculation circuit of such a ghost elimination device.

According to the operation of the filter coefficient calculating circuit of the ghost removing apparatus shown in FIG. 3, first, the field calculating unit 300 receives the received signal to remove noise from the reference signal received from the reference signal extracting circuit 12 (FIG. 1). The reference signal received by the addition of such a signal is left without reference to the reference signal modified by the characteristics of the transmission path without noise.

In FIG. 3, the correlation unit 305 takes a correlation between the signal added by the field operator 300 and the reference signal generated by the reference signal generation circuit 14 (FIG. 1). In this case, the peak value of the correlation occurs only at one position. Thereafter, the reference point setting unit 310 sets the position where the peak value of the correlation occurs in the correlation unit 305 as the reference point.

Next, a method for obtaining a filter coefficient of the FIR filter for removing the near ghost including all ghosts will be described. First, in the FIR window wing unit 32, the peak position data value obtained by the correlation unit 305 is referred to. As a result, a signal passing through the field operator 300 is taken as much as a range for removing the proximity ghost including all ghosts.

Thereafter, the FFT unit (Fast fourier transform section; 330) performs an FFT on the sample value set by the FIR window wing unit 320 to convert the data signal on the frequency axis [Gn (w)]. The calculation unit 340 divides the reference signal R (w) output from the reference signal generation circuit 14 (FIG. 1) on the frequency domain based on the data signal Gn (w). Then, the IFFT unit Inverse fourier transform section 330 may perform IFFT on the output of the first operation unit 340 to obtain a filter coefficient of the FIR filter.

Next, a method for obtaining a filter coefficient of an IIR filter for removing long ghosts will be described. First, in the IIR window wing unit 315, the long ghost is based on the data value of the peak position obtained by the correlation unit 305. FIG. The signal passing through the field operator 300 is taken as much as the range for removing the signal. After that, the FFT unit 325 performs an FFT on the sample value set by the IIR windowing unit 315 and converts the data signal G (w) on the frequency axis. Signal [G (w) multiplying signal [H1 (w)] taking FFT of signal [h1 (t)] output from IFFT unit 335 and output [G (w)] of FFT unit 325 H2 (w)] is limited in the frequency domain to the reference signal R (w) output from the reference signal generation circuit, and the third operation unit 350 subtracts the output of the second operation unit 345 from 1. Done. Next, the clamp unit 365 clamps a predetermined size or more at the output of the third operation unit 350, and the IFFT unit 370 performs an IFFT on the signal output from the clamp unit 365 to perform an IIR filter. The filter coefficient of can be obtained.

Here, according to the filter coefficient calculating circuit of the ghost elimination device, the ghost can be removed at a considerably high speed. However, if a slight error occurs in the filter coefficient due to an error occurring during the computational process for the same ghost, Since the error of filter coefficient is reflected on the screen as it is, periodic shaking phenomenon appears on the screen.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and its object is to perform a weighted average process on the filter coefficients obtained by the filter coefficient calculation circuit, and to consider the current filter coefficients and the previous weighted average continuous filter coefficients simultaneously. The present invention provides a filter coefficient weighted average processing method and a circuit of a ghost elimination device in which a screen is prevented from sudden change by adjusting a change.

According to a preferred aspect of the present invention for achieving the above object, the filter coefficient calculation circuit for calculating a filter coefficient for compensating for the distortion of the signal generated in the transmission path, the filter coefficient of the video signal input and the filter coefficient calculation circuit A ghost elimination device comprising a filter circuit for compensating for distortion generated in a transmission path based on the filter coefficient, wherein the weighted average of the FIR filter coefficients is obtained by receiving the FIR filter coefficients calculated by the filter coefficient calculation circuit. Technical means comprising an FIR filter coefficient weighted average processing unit for processing and an IIR filter coefficient weighted average processing unit for receiving the IIR filter coefficients calculated by the filter coefficient calculating circuit and performing the weighted average processing of the IIR filter coefficients are employed.

According to the present invention, each of the weighted average values unit may include: a first working memory for temporarily storing filter coefficients calculated by the filter coefficient calculating circuit; A weight setting unit that sets a weight between a previous filter coefficient and a current filter coefficient; a second working memory configured to temporarily store the weighted average processed filter coefficients last time; A weighted average calculation unit configured to calculate a current weighted averaged filter coefficient by inputting a current filter coefficient currently input, a weight of the weight setting unit, and a previous weighted averaged filter coefficient stored in the second working memory; And a control unit for controlling operations of a first working memory, the weight setting unit, the second working memory, and the weighted average calculation unit.

According to another aspect of the present invention, a filter coefficient calculating circuit for calculating a filter coefficient for compensating for distortion of a signal generated in a transmission path, and a filter coefficient of an image signal input and the filter coefficient calculating circuit are input. A ghost elimination device comprising a filter circuit for compensating for distortion generated in a transmission path as a basis, comprising: a first step of storing a current filter coefficient currently input from said filter coefficient calculating circuit in a first working memory; A second step of setting a weight between the filter coefficient and the current filter coefficient and storing the weight in the weight setting unit, the current filter coefficient and the weight and the past filter from the first working memory, the weight setting unit and the second working memory, respectively. The third step of loading the coefficient into the weighted average calculation unit, the addition of the filter coefficient from the current filter coefficients and weights and past filter coefficients A fourth step of performing an average operation, a fifth step of outputting the weighted average calculated filter coefficient to a filter circuit, and storing the weighted average calculated filter coefficient in a second working memory; and the first to fifth steps It employs a technical means consisting of repeating the steps.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

4 is a block diagram schematically showing the filter coefficient weighted average processing circuit of the ghost removing device according to the present invention.

In the figure, reference numeral 400 denotes an FIR filter coefficient weighted average processing unit 410 which receives the FIR filter coefficients calculated by the filter coefficient calculating circuit 16 and performs weighted average processing of the FIR filter coefficients, and the filter coefficient calculating circuit. A filter coefficient weighted average processing circuit comprising an IIR filter coefficient weighted average processing unit 420 that receives an IIR filter coefficient calculated in (16) and performs weighted average processing of the IIR filter coefficient.

5 is a block diagram representatively showing a specific configuration of the weighted average processing unit 410 and 420 shown in FIG. 4. In FIG. 5, reference numeral 510 denotes an FIR filter coefficient or an IIR filter calculated by the filter coefficient calculating circuit 16. First working memory for temporarily storing coefficients. 520 is a weight setting unit that sets a weight between the previous filter coefficient and the current filter coefficient. When the weight set by the weight setting unit 520 is 1α, 1α is a value between 0 and 1, approximating 1 to 1. The greater the weight of the current filter value, the greater the response to the change. The closer to 0, the smoother the change is, reflecting much of the influence of the previous filter value.

Reference numeral 530 denotes a second working memory that temporarily stores the FIR filter coefficient or IIR filter coefficient subjected to weighted average processing last time. In addition, 540 denotes a current filter coefficient (x [i]) currently input, a weight 1α of the weight setting unit 520, and a previous weighted average processed filter coefficient stored in the second working memory 530 ( A weighted average calculation unit for calculating the current weighted averaged filter coefficient (m [i]) by inputting m [i-1]), and the current weighted averaged filter coefficient m [i] is expressed by the following equation. Will be given. In other words,

m [i] = (1-α) ^* m [i-1] + α ^* x [i]

Reference numeral 500 denotes a control unit which controls the overall operation of the filter coefficient weighting unit. The operation thereof will be described with reference to FIG.

6 is an operation flowchart of the filter coefficient weighted average processing circuit shown in FIG.

First, when the filter coefficient weighted average processing circuit 400 shown in FIG. 4 starts operation, the FIR filter coefficient weighted average processing unit 410; Alternatively, the control unit 500 of the IIR filter coefficient weighted average processing unit 420 initializes the work memories 510 and 530, the weight setting unit 520, and the weighted average operation unit 540 (step 610). Subsequently, if the shape of the ghost is changed in step 615, the weighted average operation must be redone from the beginning, and it is determined whether the input signal is the same type of ghost, and if the input signal is not the same type of ghost, the process returns to step 610 On the other hand, in the case of the same type of ghost, the control proceeds to the next step.

Thereafter, the control of the control unit 500 proceeds to step 620 to determine whether the current loop is the first loop. Since the determination result at this time is an example, the control proceeds to step 630 and is currently input. The controller then stores the current filter coefficients in the first working memory 510, and then proceeds to step 640 to convert the current filter coefficients to the FIR filter 24 or the IIR filter of the filter circuit 18 as shown in FIG. In addition to outputting to (26), the current filter coefficient is stored in the second working memory 530. Then, in step 700, the current loop is increased by one, and then the flowchart shown in FIG. 6 is shown in the next filter coefficient operation. Will be executed again.

Subsequently, in step 620, since the determination result is no, the control proceeds to steps 650 and 660 to store the current filter coefficient which is currently input in the first working memory 510, and to store the previous filter coefficient between the current filter coefficients. Will set the weights.

Next, in step 670, the current filter coefficients, the weights, and the past filter coefficients are loaded from the first work memory 510, the weight setting unit 520, and the second work memory 530 into the weighted average calculation unit 540, respectively. In step 680, the weighted average operation of the filter coefficients is performed from the current filter coefficients, the weights, and the excessive filter coefficients.

Subsequently, in step 690, the weighted average calculated filter coefficient is outputted to the FIR filter 24 or the IIR filter 26 of the filter circuit 18 as shown in FIG. The filter coefficient is stored in the second working memory 530. Control then proceeds to step 700 to increment the current loop by one, and then performs the flowchart shown in FIG. 6 again in the next filter coefficient operation. Likewise, the weighted average operation of the filter coefficients is successively repeated.

As described above, a weighted average process is performed on the filter coefficients obtained by the filter coefficient calculation circuit to adjust the change of the filter coefficients in consideration of the current filter coefficients and the previous weighted average calculated filter coefficients, thereby improving the screen quality. You can do it.

Claims (3)

A filter coefficient calculating circuit 16 for calculating a filter coefficient for compensating for distortion of a signal generated in a transmission path, a video signal input, and a filter coefficient of the filter coefficient calculating circuit 16 are inputted and transmitted based on the filter coefficient. A ghost elimination device comprising a filter circuit (18) having an FIR filter (24) and an IIR filter (26) for compensating for distortion generated in a furnace, wherein the FIR filter coefficient calculated by the filter coefficient calculating circuit (16) is calculated. A weighted average processing of the FIR filter coefficients is performed and applied to the FIR filter 24 to receive the FIR filter coefficient weighted average processing unit 410 and the IIR filter coefficients calculated by the filter coefficient calculating circuit. And a filter coefficient weighting processing circuit (400) comprising an IIR filter coefficient weighted average processing unit (420) applied to the IIR filter (26) by performing a weighted average processing unit of the ghost elimination device. Emitter coefficient of weighted average processing circuit. 2. The apparatus of claim 1, wherein the weighted average processing unit (410, 420) comprises: a filter coefficient calculating circuit (16) and a first working memory (510) for temporarily storing the calculated filter coefficients; A weight setting unit 520 for setting a weight between a previous filter coefficient and a current filter coefficient; A second working memory 530 which temporarily stores the weighted average processed filter coefficients last time; A current weighted average processed filter coefficient is calculated by inputting a current filter coefficient currently input, a weight of the weight setting unit 520 and a previous weighted average processed filter coefficient stored in the second working memory 530. A weighted average operation unit 540 and; Ghost comprising a control unit 500 for controlling the operation of the first working memory 510, the weight setting unit 520, the second working memory 530, and the weighted average operation unit 540. Filter coefficient weighted average processing circuit of removal device. A filter coefficient calculating circuit 16 for calculating a filter coefficient for compensating for distortion of a signal generated in a transmission path, and an image signal input and a filter coefficient of the filter coefficient calculating circuit 16 are received and transmitted based on the filter coefficient. A ghost elimination device comprising a filter circuit (18) for compensating for distortion generated in a furnace, wherein the current filter coefficient currently being input from the filter coefficient calculating circuit (16) is stored in the first working memory (510). A first step, a second step of setting a weight between the previous filter coefficient and the current filter coefficient and storing the weight in the weight setting unit 520; the first working memory 510, the weight setting unit 520, and the first step. A third step of loading the current filter coefficients and weights and the past filter coefficients from the working memory 530 into the weighted average calculation unit 540, respectively, the addition of the filter coefficients from the current filter coefficients and weights and the past filter coefficients A fourth step of performing an average operation, a fifth step of outputting the weighted average calculated filter coefficient to the filter circuit 18 and storing the weighted average calculated filter coefficient in a second working memory 530; Filter coefficient weighted average processing method of the ghost elimination device, characterized in that the step consisting of repeating the first to fifth steps.