JPH0828064B2 - Signal interpolator - Google Patents

Description

The present invention relates to a signal interpolating device, and is particularly suitable for application when processing a digital image signal.

[Background technology and its problems]

For example, a digital video tape recorder (digital VT
In R), there is proposed an interpolator that prevents deterioration of the quality of the reproduced image by interpolating the missing sample data in the reproduced image signal based on normal sample data. As a conventional interpolation method, a method of performing an interpolation operation using data of sample points before and after adjacent to the missing sample point on a scan line including the first missing sample point is proposed. Secondly, a method of interpolating using data of upper and lower sample points included in upper and lower scan lines adjacent to the scan line including the missing sample point has been proposed. A method of performing interpolation using data at the same sample points as the sample points has been proposed.

By the way, if an attempt is made to improve the correction accuracy when the missing sample data (this is referred to as error data) is corrected by these methods, data including other errors is used for interpolation of one error data. It is generally considered desirable to use normal data for as many sample points as possible. However, when trying to use the data of a large number of sample points, an arithmetic expression that does not use the other error data is selected each time so as to adapt to the occurrence mode of the other error data (so-called adaptive). There is a problem in that, in order to actually perform this, a fairly complicated interpolation calculation must be performed.

[Object of the Invention]

The present invention has been made in consideration of the above points, and it is an object of the present invention to propose a signal interpolating device which has a relatively simple configuration in order to realize adaptive interpolation calculation with high modification accuracy.

[Outline of Invention]

In order to achieve such an object, in the present invention, when an error occurs in a digital input image signal formed by sequentially transmitting sample data at a predetermined time interval, the error sampling on the screen is performed by a first-stage interpolation circuit. A corrected image signal corrected by the first-stage interpolation circuit by interpolating error sample points based on normal sample data of samples facing the point and sandwiching the error sample points. To the second-stage interpolating circuit, so that the first interpolated data at the error sampling point is subjected to error sampling so that the change on the time axis of the first interpolated data and the sample data before and after the first interpolated data becomes smooth. Signal interpolation that interpolates points to obtain the second interpolation data, which makes it possible to perform interpolation calculation with high correction accuracy with a relatively simple configuration. To be able to get location.

〔Example〕

An embodiment of the present invention will be described in detail below with reference to the drawings. First
The figure shows a signal interpolating apparatus as a whole, and a first stage interpolating circuit 1 for receiving an input video signal VD to be interpolated and a first stage interpolating circuit S1 for receiving an interpolating output signal S1 obtained by the first stage interpolating circuit 1 It has a two-stage interpolation circuit 3,
The output S2 of the second stage interpolation circuit 3 is sent out as a data output signal S3 through the switch circuit 4.

When an error occurs in the input video signal VD, the first-stage interpolator 1 correlates the data D _{nm of the} sample point where the error occurs with the data of the sample points around it as shown in FIG. It is designed to perform low-precision interpolation calculation using data having a high property. Regarding the scan line L _n including the sample point where an error has occurred on the screen, data D _{n (m-1)} and D _{n (m + 1) of} the sample points adjacent to the error point are laterally present in the scan line L _n. Exists. The line L _(n-1) adjacent to and below the line L _n containing the error point and L
Data D _{(n-1) m} of vertically adjacent sample points on _{(n + 1)}
And D _{(n + 1) m} exist. In addition to this, data D _{(n-1) (m-1)} , D _{(n + 1) (m + 1)} and
There are D _{(n-1) (m + 1)} and D _{(n + 1) (m-1)} .

Therefore, the data that can be used in the predictive calculation of the data D _nm of the error sample points are the sample points facing the error sample points and sandwiching them, that is, two points in the horizontal direction,
There are two points in the vertical direction, two points in the diagonal direction from the upper left to the lower right, and two points in the diagonal direction from the upper right to the lower left, and the correlation of the data of the adjacent sample points in each direction can be seen.

The first-stage interpolator 1 judges the vertical and horizontal correlations among these four-direction correlations and has a strong correlation in the horizontal direction (that is, the horizontal direction), or a vertical direction (that is, the vertical direction). Of the error point data D _nm is determined by using the data in the direction in which the correlation is strong. That is, the first stage interpolation circuit 1 separates the input video signal VD into the luminance signal Y and the chroma signal C in the luminance signal / chroma signal separation circuit 11 and inputs the luminance signal Y and the chroma signal C to the horizontal direction interpolation circuit 13. Also, 1 for the input video signal VD
Similarly, the video signal VD _n-1 before the line is separated into the brightness signal and the chroma signal C in the brightness signal / chroma signal separation circuit 14 and is then applied to the vertical direction interpolation circuit 12, and similarly to the input video signal VD. On the other hand, the video signal VD _{n + 1} after 1H is separated into the luminance signal Y and the chroma signal C in the luminance signal / chroma signal separation circuit 15 and applied to the vertical direction interpolation circuit 12.

The vertical and horizontal interpolation circuits 12 and 13 are provided with position information EI about the sample points where an error has occurred, so that in the vertical direction interpolation circuit 12, they are located above and below the sample points where an error has occurred. Sample point data D
_{Based on (n-1) m} and D _{(n + 1) m} , interpolation calculation is performed on the data D _nm of the sample point where the error has occurred, and the interpolation calculation output S11
To the switching switch circuit 16. The horizontal interpolation circuit
Reference numeral 13 is an interpolation calculation based on the lateral data D _{n (m-1)} and D _{n (m + 1)} sandwiching an error point designated based on the error information EI, and the interpolation calculation output S12 is used as a switch circuit 16 Send to.

The luminance signals Y of the separation circuits 14 and 15 are given to the vertical direction correlation determination circuit 17 to determine the correlation strength of the corresponding sample points for the lines around 1H and send the determination output S13 to the comparison / selection circuit 18. . On the other hand, the luminance signal Y of the luminance signal / chroma signal separation circuit 11 is given to the horizontal direction correlation determination circuit 19, whereby the correlation strength of the data at the adjacent sample points on the line L _n is set. After making a determination, the determination output S14 is sent to the comparison / selection circuit 18. Comparison selection circuit 18
Compares the strength of the correlation represented by the discrimination outputs S13 and S14, and sends the switching control signal S15 for switching the switch circuit 16 to the horizontal interpolation circuit 13 side when the stronger correlation is in the horizontal direction. On the other hand, if the one having a stronger correlation is in the vertical direction, the switching control signal S15 for switching the switch circuit 16 to the vertical direction interpolation circuit 12 side is transmitted.

Thus, the first-stage interpolation circuit 1 selects the data in the direction having strong correlation among the adjacent data in the vertical direction or the horizontal direction around the error point designated by the error information EI and executes the interpolation calculation. Probable interpolation calculation data is obtained as missing data D _nm .

In the above description, the case where the interpolation calculation is executed in the horizontal direction and the vertical direction has been described, but in addition to or instead of this, the interpolation calculation is performed using the data in the diagonal direction as described above with reference to FIG. May be configured to execute.

Further, in the case of FIG. 3, the case where the interpolation data is obtained based on the composite signal has been described, but when processing the component signal, the same configuration as that in which the luminance signal / chroma signal separation circuits 11, 14, 15 are omitted The first stage interpolation circuit 1 may be provided.

As described above, the first-stage interpolation circuit 1 executes the interpolation operation using only the data adjacent to the sample point where the error has occurred. Therefore, the degree of correction of the data is low in precision, but is relatively simple. With this configuration, it is possible to interpolate the data most likely to have the correlation, and the data is sent to the second-stage interpolating circuit 3 via the switch circuit 2 as the interpolating output S1.

The second-stage interpolation circuit 3 executes an interpolation operation based on the input video signal VD in which the interpolation data is inserted so that the smoothness is not lost due to the data change in the horizontal direction due to the insertion of the interpolation data. Therefore, high-accuracy retouching data is obtained, and for example, a digital lopal filter or an interpolation filter by polynomial approximation is used.

For example, when an interpolation filter is used, the second-stage interpolation circuit 3 having the configuration shown in FIG. 4 can be applied. In this configuration, the video signal including the data x ₀ of the sample point where the error has occurred is arranged as x ₁ , x ₂ ... in the direction toward the past and x in the direction toward the future as shown in FIG. Assuming that the data is arranged as _-1 , x _-2 …, the value that does not lose smoothness in view of the change in the time axis of the data connecting the surrounding data as the data x ₀ of the sample point where the error occurred Interpolate x ₀ .
Therefore, when the data x ₀ is obtained on condition that the n-th order difference is equal, x ₀ = a ₁ (x ₁ + x _-1 ) + a ₂ (x ₂ + x _-2 ) + ... + a _n (x _n + x _-n ) ……
It suffices to execute the calculation of (1).

In the case of FIG. 4, it is possible to obtain interpolated data with a practically sufficient correction accuracy by taking the fourth-order difference. In this case, the interpolated data x ₀ is x ₀ = a ₁ (x ₁ + x _-1 ) ＋ a ₂ (x ₂ + x _-2 ) ＋ a ₃ (x ₃ + x _-3 ) ＋ a ₄ (x ₄ + x _-4 )
...... It is composed of a digital filter that calculates the interpolation calculation formula of (2). That is, the sample data sequentially arriving in each sample section T of the video signal interpolated by the first-stage interpolation circuit 1 are transmitted through the delay circuits 31, 32, ... Go. Therefore, when the error point data x ₀ is output to the output terminal of the central delay circuit 35, the data obtained at the output terminals of the delay circuits 36, 37, 38, 39 provided in the data transmission direction. Is the data x ₁ , x ₂ , x ₃ , x ₄ at the sampling points in the past direction, and the output terminals of the delay circuits 34, 33, 32, 31 provided in the opposite direction to the transmission direction. The data obtained in the direction of the future is x _-1 , x _-2 , x _-3 , x _-4 .

Then, the data x ₁ and x _-1 , x ₂ and the data obtained by shifting by 1 sample point in the past and future directions with reference to the data x ₀ of the sample point where the error occurred
x _-2 , x ₃ and x _-3 , x ₄ and x _-4 are adder circuits 41, 42, ₄ respectively.
3, 44, and the added output is added to the coefficient circuits 51, 52,
After being multiplied by a ₁ , a ₂ , a ₃ , and a ₄ in 53 and 54, respectively, they are added in an adding circuit 61. Thus, the interpolation calculation data S2 having the contents of the equation (2) is sent from the adder circuit 61.

Thus interpolation operation results obtained by the first-stage interpolator data x ₁ required compared greatly manner but modification accuracy is obtained as a high data according calculated and computed interpolation data in 1 and x _{- 1} ~x ₄ and x _-4 because no missing can be performed reliably (when missing is interpolated by being in the first stage interpolator 1) according interpolation. Because of this, the error point data x ₀ that you are trying to interpolate now
Is obtained by the interpolation calculation in the first-stage interpolation circuit 1 though it is rough, and the past and future data sandwiching this data x ₀ include an error point (the input video signal VD (In the step), since this error point is interpolated by the interpolation operation of the first stage interpolation circuit 1, all the data necessary for the calculation in the configuration of FIG. 4 will be supplied. The data of the past and future sample points used for the interpolation calculation is obtained as the data having the strongest correlation with the data of the normal sample points in the first-stage interpolation circuit 1 as a reference. It can be said that the content of the interpolation calculation output S2 obtained at the output end of the interpolation circuit 3 has a highly accurate modification content.

As described above, according to the configuration of FIG. 1, the second-stage interpolation circuit 3 executes the high-precision interpolation calculation based on the data obtained by the low-precision interpolation calculation in the first-stage interpolation circuit 1. By doing so, a data signal having a high modification accuracy of the second stage interpolation circuit 3 can be obtained as the output data output signal S3. On the other hand, when the input video signal VD does not include an error sample point, the input video signal VD passes through the delay circuit 65 (the delay time of which corresponds to the delay time of the first stage interpolation circuit 1) and the switch circuit. 2 is provided to the delay circuit from the switch circuit 2.
It is given to the switch circuit 4 through 66 (the delay time thereof corresponds to the second stage interpolation circuit 3). The error information EI is provided to the switch circuits 2 and 4 through the delay circuits 67 and 68 after being adjusted so that the timing is adjusted, and thus the switch circuits 2 and 4 are interpolated at the first stage at the timing of the sample point where the error occurs. The circuit 1 and the second-stage interpolation circuit 3 are switched to each other, and the switch circuits 2 and 4 are switched to the delay circuits 65 and 66 at the timing of the sampling point where no error occurs. Thus, when an error occurs, the interpolation calculation for the sample point is executed, and when the error does not occur, the input video signal VD is sent as it is as the data output signal S3.

〔The invention's effect〕

As described above, according to the present invention, based on the digital data obtained by performing the interpolation calculation using the data of the two sample points which face the error sample point and sandwich the error sample point in the first stage interpolation circuit 1, The second-stage interpolation circuit 3 recalculates the interpolation data using the data of the past and future sample points including the error point so that the change does not lose smoothness. The data output signal S3 having a high retouching accuracy of 3 can be reliably obtained. First stage interpolation circuit 1
Has an adaptive function so that it can adapt to the array of sample points of the error that has occurred, but it can be applied with a low correction accuracy, so that its configuration is actually simple. Therefore, it is possible to realize a signal interpolating device having a remarkably simple structure as a whole as compared with the case where a large number of normal sample points have to be extracted in order to improve the correction accuracy based on the conventional idea.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the signal interpolating device according to the present invention, FIG. 2 is a schematic diagram showing the operating principle of the first stage interpolating circuit, and FIG. 3 is the first stage interpolating circuit of FIG. FIG. 4 is a block diagram showing a detailed configuration of the circuit 1, FIG. 4 is a block diagram showing a detailed configuration of the second stage interpolation circuit 3, and FIG. 5 is a signal wave system diagram for explaining the operating principle thereof. 1 ... 1st stage interpolation circuit, 2, 4 ... Switch circuit, 3 ...
Second stage interpolation circuit, 65-68 ... Delay circuit.

Claims (1)

[Claims] 1. When an error occurs in a digital input image signal which is obtained by sequentially transmitting sample data at a predetermined time interval, a sampling point on the screen facing the error sampling point and sandwiching the error sampling point is displayed. A first-stage interpolation circuit for interpolating the error sample points to obtain first interpolation data based on a pair of sample data having the highest correlation among a pair of normal sample data; and a first-stage interpolation circuit. For the first interpolation data of the error sample point that has received the modified image signal thus corrected, the change on the time axis of the first interpolation data and the sample data of the sample points before and after the error interpolation point are smoothed. A second-stage interpolation circuit for interpolating the error sample points to obtain second interpolation data.