CN1096185C - Interlaced-to-progressive conversion apparatus and method using motion and spatial correlation - Google Patents

Abstract

The interlaced to progressive conversion device includes a spatial interpolator, which spatially interpolates the input interlaced image signal and outputs a spatially interpolated signal; a time interpolator, which temporally interpolates the input interlaced image signal and outputs a time interpolated signal; a correlator , to detect the motion correlation, the vertical correlation, and the time vertical correlation by using a predetermined number of sampled data in the current field, the previous field, and the next field; and a selector for comparing the motion correlation, the vertical correlation, and the time vertical correlation A predetermined constant is compared, and one is selected between the spatially interpolated signal and the temporally interpolated signal according to the comparison result. It can enhance the reliability of motion information and reduce artifacts by temporally or spatially interpolating interlaced image signals using motion and spatial correlation.

Description

Apparatus and method for interlaced to progressive conversion using motion and space correlation

The present invention relates to an interlaced-to-progressive conversion apparatus and method using motion and space correlation. In particular, the present invention relates to an apparatus and method for converting an interlaced image signal into a progressive image signal by spatially or temporally interpolating depending on motion and spatial correlation.

Generally, interlace-to-progressive conversion (IPC) devices have been widely used to reduce many artifacts caused by interlacing, namely, vertical resolution degradation, scan line flicker, and wide area flashes.

Recently, interlace-to-progressive converters have become more important because high-definition television (HDTV) systems employ multiple standards for signal formats and often require conversion between standard input/output signals of different formats.

The interlaced to progressive conversion algorithm that has been developed at an early stage is used by systems such as NTSC, PAL, SECAM, etc. Algorithms that have been proposed are generally based on the interpolation of replaying lines excluded from interlacing.

IEEE Transactions on Consumer Electronics, Vol. 35, No. 3, pp. 279-289, Aug. 1989, by D.I.C. Hentschei entitled "Interpolation for Median Filtering and Vertical Edge Control for Flicker Reduction The proposed algorithm to reduce these artifacts, an interpolation method based on simple line repetition and vertical filtering, and a method based on vertical edge information are disclosed in reference [1] of "Comparison of Imagers".

In reference [2] entitled "HD-TV Space-Time Upconversion" by D. Bagni, R. Lancini, S. Landi, and S. Tubaro, Proceedings of the International Symposium on HDTV, 1994 Interpolation methods according to edge directions are disclosed.

In May 1989, IEEE ISCAS-89 Proceedings, Portland, USA, pp. 433-436, published by J. Juhola, A. Nieminen, J. Salo, and Y. Neuvo entitled "Scan Rate Conversion Using Weighted Median Filtering "A non-linear high-speed interpolation method based on a weighted median filter is disclosed in the reference [3].

References entitled "Nonlinear Five-point Interpolation Filtering" published by A. Lehtonen and M. Renfors in "SPIE Visual Communication and Image Processing Transactions", Lausanne, Switzerland, October 1990, pp. 132-135 [ 4] discloses an algorithm based on the FIR median hybrid filter.

Reference by H. Blume, L. Schwoerer, and K. Zygis in Proceedings of the International Symposium on HDTV, 1994 entitled "Subbands based on up-conversion using a complementary median filter" [5] An algorithm based on a complementary median filter is disclosed in .

Reference by T. Doyle entitled "Interlaced to Progressive Conversion for EDTV Applications", L. Chiariglione Ed., North Holland, Elsevier Scientific Publishers, HDTV Signal Processing, 1988, pp. 421-430 An algorithm based on orientation-dependent median filtering is disclosed in [6].

References entitled "Two Integrated Progressive Scan Converters" published by P. Frenchn in IEEE Transactions on Consumer Electronics, Vol. 32, No. 3, pp. 237-240, 1986 [7]; and " IEEE Technical Paper Abstracts, pp. 186-187, by T. Doyle and P. Frencken, reference [8] entitled "Median Filtering of Television Images", discloses an algorithm based on a vertical-temporal median filter.

Also, in IEEE Transactions on Consumer Electronics, Vol. 33, No. 3, pp. 266-271, August 1989, by N. Suzuki et al., entitled "Use in Perfecting Motion-Adaptive Pre-Scan Conversion in IDTV Receivers Improved Synthetic Motion Signals" in reference [9]; and IEEE Transactions on Consumer Electronics, Vol. 36, No. 2, pp. 110-114, May 1990, by C.P. Markhauser entitled "With Two-Dimensional Profiles Enhanced Motion-Adaptive Pre-Scan Converter" reference [10] discloses a motion-adaptation scheme.

The above interlaced to progressive conversion methods can be roughly divided into spatial interpolation methods, temporal interpolation methods, and three-dimensional interpolation methods combining spatial interpolation and temporal interpolation.

In the three-dimensional interpolation method, since erroneous temporal interpolation can cause image quality degradation such as tearing-artifact, detecting motion in an image and performing temporal interpolation appropriately according to the detected motion are very important. This is due to the fact that the theoretically determinable maximum temporal frequency is limited due to the fact that the temporal sampling rate in actual image signals is smaller than the Nyquist rate.

However, as proposed by the present invention, the reliability of the motion information can be increased by using the sampled spatial information together with the motion information.

An object of the present invention is an interlaced-to-progressive conversion device for converting an interlaced image signal into a progressive image signal by performing spatial interpolation or temporal interpolation according to motion and spatial correlation.

Another object of the present invention is a three-dimensional interlaced-to-progressive conversion method for selecting and outputting one of spatially interpolated or temporally interpolated image signals depending on motion and spatial correlation.

In order to achieve the above object, there is provided an interlaced-to-progressive conversion device for converting an input interlaced image signal into a progressive image signal, comprising: a spatial interpolation device for spatially interpolating the input interlaced image signal and outputting Spatial interpolation signal; time interpolation means for time interpolating the input interlaced image signal and outputting time interpolation signal; correlation means for using a predetermined number of samples in the current field, previous field and next field data to input the input interlaced image signal and output motion correlation, vertical correlation, and time vertical correlation; and selecting means for comparing motion correlation, vertical correlation, and time vertical correlation with corresponding predetermined constants, and according to the comparison As a result, one is selected between a spatially interpolated signal and a temporally interpolated signal.

In order to achieve the above another object, there is provided a method for converting an input interlaced image signal into a progressive image signal from interlaced to progressive, comprising the steps of: (a) spatially interpolating the input interlaced image signal and outputting a spatially interpolation signal; (b) time interpolating the input interlaced image signal and outputting the time interpolation signal; (c) detecting motion correlation, vertical correlation, and time vertical correlation from the input interlaced image signal; (d) if The detected motion correlation value is greater than the first predetermined constant, and the space interpolation signal is selected; (e) if the motion correlation value is not greater than the first predetermined constant and the detected vertical correlation value is greater than the second predetermined constant, then the time interpolation signal is selected (f) if the motion correlation value is not greater than a first predetermined constant and the vertical correlation value is not greater than a second predetermined constant and the detected time vertical correlation value is greater than a third predetermined constant, then select the spatially interpolated signal; and (g) if The motion correlation value is not greater than a first predetermined constant and the vertical correlation value is not greater than a second predetermined constant and the time vertical correlation value is not greater than a third predetermined constant, and the signal is interpolated over time.

The above objects and advantages of the present invention will become more apparent by describing in detail preferred embodiments thereof with reference to the accompanying drawings. in:

Fig. 1 is a block diagram according to an embodiment of interlaced to progressive conversion device of the present invention;

Fig. 2 is a detailed block diagram of the motion spatial correlator shown in Fig. 1;

Figure 3 illustrates the geometry of the samples used to switch the spatially or temporally interpolated signal; and

FIG. 4 is a flow chart describing the method for switching the spatial or temporal interpolation signal by the motion spatial correlator in the selector shown in FIG. 1 according to the output signal.

In the following, an apparatus and method for interlaced to progressive conversion using motion and space correlation and its preferred embodiments will be described with reference to the accompanying drawings.

Fig. 1 is a block diagram of an embodiment of an interlaced to progressive conversion device according to the present invention.

The interlaced to progressive converting device includes a spatial interpolator 110 , a temporal interpolator 120 , a correlator 130 and a selector 140 .

The spatial interpolator 110 spatially interpolates the input interlaced image signal Vin using a predetermined algorithm.

The time interpolator 120 time interpolates the input interlaced image signal Vin using a predetermined algorithm.

The correlator 130 detects a motion correlation DM, a vertical direction correlation (hereinafter referred to as "vertical direction correlation") DV, and a correlation between temporal interpolation and vertical interpolation (hereinafter referred to as "temporal correlation") from the input interlaced image signal. vertical correlation") DT.

The selector 140 compares the motion correlation DM, the vertical correlation DV, and the time vertical correlation DT with predetermined constants TM, T1, and T2 respectively, and selects the signal Is output by the spatial interpolator 110 or the signal Is output by the temporal interpolator according to the comparison result. The signal It output by 120 can be output as the interpolation signal Vout.

The operation of the device shown in Fig. 1 will now be described.

In FIG. 1 , an interlaced image signal Vin is applied to a spatial interpolator 110 , a temporal interpolator 120 , and a correlator 130 .

Meanwhile, since the present invention involves switching between a spatially interpolated progressive image signal (hereinafter referred to as "spatial interpolated signal") and a temporally interpolated progressive image signal (hereinafter referred to as "temporal interpolated signal") based on motion information and spatial information, interpolation signal") and output the selected signal, so it does not matter which spatial interpolation algorithm and temporal interpolation algorithm are used in the spatial interpolator 110 and temporal interpolator 120 respectively.

The correlator 130 inputs the interpolation signal Vin and detects motion correlation DM, vertical direction correlation DV, and time vertical correlation DT from the interlaced image signal Vin. Thereafter, the correlator 130 outputs signals DM, DV and DT.

The selector 140 compares the motion correlation DM, the vertical correlation DV, and the time vertical correlation DT with predetermined constants TM, T1, and T2, respectively, and selects the signal Is output by the spatial interpolator 110 or the signal Is output by the temporal interpolator 120. Signal It, and output the selected signal Vout.

FIG. 2 is a detailed block diagram of the correlator shown in FIG. 1. FIG.

The sample delays 201, 203, 204, and 206, and the row delays 202 and 205 constitute a first sample detector, and detect samples w1 to w5. The field memory 210, sample delayers 211 and 213, and row delayer 212 constitute a second sample detector, and detect samples x1 and x2. In addition, the field memory 220, the sample delayers 221, 223, 224, and 226, and the row delayers 222 and 225 constitute a third sample detector, and detect samples v1 to v5.

The motion correlator 230 receives the samples v1 and v5 and w1 and w5, and calculates and outputs the motion correlation DM. The subtractor 241 and the absolute value circuit 242 receive the samples v1 to v3 and w1 to w5, and calculate and output the vertical correlation DV. In addition, the adders 251 and 252, the subtractor 253 and the absolute value circuit 254 receive the samples x1, x2, v3 and w3, and calculate and output the time vertical correlation DT.

FIG. 3 illustrates the geometric relationship between the samples shown in FIG. 2. FIG.

In FIG. 3, sample "x" represents the sample to be recovered by interpolation (referred to as the "current interpolated sample"). A sample "x1" indicates a sample in the previous row having the same horizontal position as the current interpolated sample. In addition, the sample "x2" indicates the sample in the next row having the same horizontal position as the current interpolation sample.

Sample "v3" indicates the sample in the previous field having the same position as the current interpolated sample. Samples "v2" and "v4" are the previous and subsequent samples of sample v3. In addition, a sample "v1'" indicates a sample having the same horizontal position as the sample v3 in the previous line. A sample "v5" indicates a sample having the same horizontal position as the sample v3 in the next line.

Likewise, sample "w3" indicates a sample in the next field having the same position as the current interpolated sample. The samples "w2" and "w4" are the previous and subsequent samples of the sample w3. In addition, the sample "w1" indicates a sample having the same horizontal position as the sample w3 in the previous line. Sample "w5" indicates a sample in the next row having the same horizontal position as sample w3.

Likewise, if the current field is an odd field, since the input signal is an input interlaced signal, the previous field and the next field are even fields. In addition, if the current field is an even field, the previous field and the next field are odd fields.

Referring now to FIG. 3, the operation of the apparatus shown in FIG. 2 will be described.

In FIG. 2 , an interlaced image signal Vin is input to a field memory 210 , a sampling delayer 201 and a line delayer 202 .

The field memory 210 stores one field of the input image signal Vin and outputs the image signal delayed by one field period. The field memory 220 inputs the delayed image signal from the field memory 210, stores one field of the delayed image signal, and outputs the image signal delayed by two field periods compared with the input original input signal Vin. The field memories 210 and 220 are composed of first-in-first-out (FIFO) memories.

The sampling delayer 201 inputs the interlaced image signal Vin, delays the signal by one sampling period, and outputs the sampling w5 of the next field shown in FIG. 3 . The line delayer 202 receives the input signal Vin, delays the signal by one horizontal line period, and outputs the sample w4 when the sample delayer 201 outputs the sample w5. The sampling delayer 203 delays the signal output by the horizontal delayer 202 by one sampling period, and outputs a sampling signal w3. The sampling delayer 204 delays the signal output by the sampling delayer 203 by 1 sampling period, and outputs the sampling signal w2. The line delayer 205 delays the signal output from the line delayer 202 by one horizontal line period. The sampling delayer 206 delays the signal output by the horizontal delayer 205 by one sampling period, and outputs a sampling signal w1.

Meanwhile, the sample delayer 211 further delays the signal delayed by 1 field period from the field memory 210 by 1 sample period, and outputs sample x2. The line delayer 212 delays the signal output from the field memory 210 by one horizontal line period. The sampling delayer 213 delays the signal output by the horizontal delayer 212 by one sampling period, and outputs the sampling signal x1.

The sample delayer 221 further delays the signal delayed by 2 field periods from the field memory 210 by 1 sample period, and outputs a sample v5. The line delayer 222 delays the signal output from the field memory 220 by one horizontal line period, and outputs a sample v4. The sampling delayer 223 delays the signal output by the horizontal delayer 222 by 1 sampling period, and outputs a sampling signal v3. The sampling delayer 224 delays the signal output by the sampling delayer 223 by 1 sampling period, and outputs a sampling signal v2. The line delayer 225 delays the signal output from the line delayer 222 by one horizontal line period. The sampling delayer 226 delays the signal output by the horizontal delayer 225 by 1 sampling period, and outputs a sampling signal v1.

The motion correlator 230 receives the samples v1 to v5 of the previous field and the samples w1 to w5 of the next field, calculates and outputs the motion correlation DM. At this time, the motion correlation DM output by the motion correlator 230 is calculated by formula (1). $\mathrm{DM}=\underset{i=1}{\overset{5}{\mathrm{\Σ}}}\mathrm{ai}|\mathrm{vi}-\mathrm{wi}|...\left(1\right)$ Among them, ai is a preset coefficient.

Motion-related DM is a measurement used to estimate whether motion is generated adjacent to the position of the sample to be recovered by interpolation (x in FIG. 3 is the geometric midpoint of x1 and x2).

The subtracter 241 subtracts the sampled data x2 output from the sampling delayer 211 from the sampled data x1 outputted from the sampling delayer 213, and outputs the subtraction result. Thereafter, the absolute value circuit 242 calculates the absolute value of the subtraction result and outputs the absolute value as the vertical correlation DV.

Therefore, formula (2) can be used to express the vertical correlation DV.

...(2)

The adder 251 adds the sample data v3 to the sample data w3. The adder 252 adds the sampled data x1 to the sampled data x2.

The subtracter 253 subtracts the signal output from the adder 252 from the signal output from the adder 251 and outputs the subtraction result. Thereafter, the absolute value circuit 254 calculates the absolute value of the subtraction result and outputs the absolute value as the time vertical correlation DT.

Therefore, formula (3) can be used to express the time vertical correlation DT.

                                                                                                                                      Time

The motion correlation DM output from the motion correlator 230 , the vertical correlation DV output from the absolute value circuit 242 , and the time vertical correlation DT output from the absolute value circuit 254 are input to the selector 140 shown in FIG. 1 .

The operation of the selector 140 is described with reference to the flow shown in FIG. 4 .

In FIG. 4, the selector 140 compares the motion-related DM with a constant TM (step S101). If the motion correlation DM is larger than the constant TM indicating that motion is generated, the selector 140 selects and outputs the spatially interpolated signal from the spatial interpolator 110 (step S102).

If the motion correlation DM is not greater than the motion constant TM in step S101, the selector 140 compares the vertical correlation DV with the constant T1 (step S103). If the vertical correlation DV is greater than the constant T1, the selector 140 selects and outputs the time-interpolated signal It from the time-interpolator 120 (step S104).

Here, the reason why the vertical correlation DV is compared with the constant T1 when the motion correlation DM is not greater than the constant TM is because artifacts that may be generated by erroneous temporal interpolation are visually perceived differently according to the magnitude of the vertical correlation DV. When the vertical correlation DV is greater than the constant T1, since the image signal is less correlated in the vertical direction, it is not easy to detect the artifacts caused by wrong time interpolation visually. Therefore, when DM is not greater than TM and DV is greater than T1, the signal It from the time interpolator 120 is selected and output by the selector 140 (step S104).

However, when DM is smaller than TM and DV is smaller than T1, artifacts due to erroneous temporal interpolation are more easily perceived visually. Therefore, the selector 140 selects and outputs the signal Is from the spatial interpolator 110 or the signal It from the temporal interpolator 120 according to the temporal vertical correlation DT indicating the correlation between temporal interpolation and vertical interpolation.

That is, when the motion correlation DM is not larger than the constant TM and the vertical correlation DV is not larger than the constant T1, the selector 140 compares the time vertical correlation DT with the constant T2 (step S105). If the time vertical correlation DT is larger than the constant T2, since the correlation between the time interpolation and the vertical interpolation is small, the selector 140 selects and outputs the signal Is from the spatial interpolator 110, so as to reduce the influence of artifacts (step S106 ). Meanwhile, if the time vertical correlation DT is not greater than the constant T2, the selector 140 selects and outputs the output signal It from the time interpolator 120 because the correlation between time interpolation and vertical interpolation is large (step S107).

As described above, the interlaced-to-progressive converting apparatus and method according to the present invention temporally or spatially interpolate an interlaced image signal using both motion and spatial correlation, and enhance reliability of motion-related information and effectively reduce artifacts.

Claims (14)

1, a kind of be used for will input the interlaced image signal interlacing that converts progressive video signal to conversion equipment line by line, comprising:

The spatial interpolation device is used for interlaced image signal and output region interpolated signal that spatial interpolation is imported;

The temporal interpolation device is used for interlaced image signal and output time interpolated signal that temporal interpolation is imported;

Relevant apparatus is used for interlaced image signal and relevant, vertical relevant be correlated with vertical with the time of output movement imported by using the sampled data of working as front court, previous field and next predetermined quantity to import; With

Choice device is used for relevant, the vertical relevant vertical relevant and corresponding predetermined constant with the time of motion relatively, and selects one according to comparative result between spatial interpolation signal and temporal interpolation signal.

2, interlacing according to claim 1 is to conversion equipment line by line, and wherein said space relevant apparatus comprises:

First checkout gear, the motion that is used to detect between the first sampling group and the second sampling group is relevant, this first sampling group is comprising locational first sampling identical with current interpolation sampling, sampling after the sampling before first sampling and first sampling and in previous row and back delegation with first sampling in the previous field of the locational sampling of par, this second sampling group is comprising locational second sampling identical with the sampling of current interpolation, second the sampling before sampling and second the sampling after sampling and previous row and the back delegation in second the sampling the locational sampling of par next in;

Second checkout gear is used for coming detection of vertical relevant by calculating with the difference of current interpolation sampling between the sampling of locational previous row of working as the front court of par and next line; With

The 3rd checkout gear, be used for by calculate first sampling and second sampling and with par locational when the front court previous row and the sampling of next line and between difference come detection time vertical relevant.

3, interlacing according to claim 2 is to conversion equipment line by line, and wherein said second checkout gear comprises:

First subtracter is used for deducting the sampling of par position in the current delegation after the match and exporting first and subtract each other the result from the sampling when the previous row par position, front court; With

First absolute value circuit is used for determining that first subtracts each other result's absolute value.

4, interlacing according to claim 2 is to conversion equipment line by line, and wherein said the 3rd checkout gear comprises:

First adder is used for first sampling is added to second sampling;

Second adder is used for when the previous row and the locational sampling addition of next line par of front court;

Second subtracter is used for deducting the output of described second adder and exporting second and subtract each other the result from the output of described first adder; With

Second absolute value circuit is used for definite second absolute value and the output time that subtracts each other the result and vertically is correlated with.

5, interlacing according to claim 2 further comprises to conversion equipment line by line:

First sampling detecting device is used to detect the first sampling group;

Second sampling detecting device is used for detecting with current interpolation sampling locational when the previous row of front court and the sampling of next line at par; With

The 3rd sampling detecting device is used to detect the second sampling group.

6, interlacing according to claim 5 is to conversion equipment line by line, and wherein said first sampling detecting device comprises:

The first sampling delayer is used to import interlaced image signal also with sampling period of this signal delay;

The first row delayer is used to import interlaced image signal also with horizontal line cycle of this signal delay;

The second sampling delayer, be used to import the described first row delayer output and should the sampling period with one of this signal delay;

The 3rd sampling delayer, be used to import the described second sampling delayer output and should the sampling period with one of this signal delay;

The second row delayer is used to import the output of the described first row delayer and with this horizontal line cycle of this signal delay; With

The 4th sampling delayer, be used to import the described second row delayer output and should the sampling period with one of this signal delay.

7, interlacing according to claim 5 is to conversion equipment line by line, and wherein said second sampling detecting device comprises:

First field memory is used to import interlaced image signal also with field duration of this signal delay;

The 5th sampling delayer, be used to import described first field memory output and should the sampling period with one of this signal delay;

The third line delayer is used to import the output of described first field memory and with this horizontal line cycle of this signal delay; With

The 6th sampling delayer, be used to import described the third line delayer output and should the sampling period with one of this signal delay.

8, interlacing according to claim 5 is to conversion equipment line by line, and wherein said the 3rd sampling detecting device comprises:

Second field memory, be used to import described first field memory output and should the field duration with one of this signal delay;

The 7th sampling delayer, be used to import described second field memory output and should the sampling period with one of this signal delay;

The fourth line delayer is used to import the output of described second field memory and with this horizontal line cycle of this signal delay;

The 8th sampling delayer, be used to import described fourth line delayer output and should the sampling period with one of this signal delay;

The 9th sampling delayer, be used to import described the 8th sampling delayer output and should the sampling period with one of this signal delay;

The fifth line delayer is used to import the output of described fourth line delayer and with this horizontal line cycle of this signal delay; With

The tenth sampling delayer, be used to import described fifth line delayer output and should the sampling period with one of this signal delay.

9, interlacing according to claim 1 is to conversion equipment line by line, if wherein the motion correlation is greater than one first predetermined constant, described choice device is selected and the output region interpolated signal.

10, interlacing according to claim 1 is to conversion equipment line by line, if wherein the motion correlation be not more than this first predetermined constant and vertical correlation greater than one second predetermined constant, described choice device is then selected and the output time interpolated signal.

11, interlacing according to claim 1 is to conversion equipment line by line, if wherein the motion correlation is not more than this first predetermined constant and is not more than this second predetermined constant with vertical correlation, and vertical correlation of the time of being detected is greater than one the 3rd predetermined constant, and described choice device is then selected and the output region interpolated signal.

12, interlacing according to claim 1 is to conversion equipment line by line, if wherein the motion correlation is not more than that this first predetermined constant is not more than this second predetermined constant with vertical correlation and vertical correlation of time is not more than the 3rd predetermined constant, described choice device is then selected and the output time interpolated signal.

13, a kind of be used for will input the interlaced image signal interlacing that converts progressive video signal to conversion method line by line, may further comprise the steps:

(a) interlaced image signal and the output region interpolated signal imported of spatial interpolation;

(b) interlaced image signal and the output time interpolated signal imported of temporal interpolation;

(c) it is relevant, vertical relevant vertical with the time relevant to detect motion from the interlaced image signal of being imported;

(d) if the motion correlation that is detected greater than one first predetermined constant, is then selected the spatial interpolation signal;

(e) if the motion correlation is not more than this first predetermined constant and the vertical correlation that detected greater than one second predetermined constant, select time interpolated signal then;

(f) if the motion correlation is not more than that this first predetermined constant is not more than this second predetermined constant with vertical correlation and the vertical correlation of time that detected greater than one the 3rd predetermined constant, select time interpolated signal then; With

(g) this first predetermined constant is not more than this second predetermined constant with vertical correlation and vertical correlation of time is not more than the 3rd predetermined constant if the motion correlation is not more than, then the select time interpolated signal.

14, interlacing according to claim 13 is to conversion method line by line, and wherein said step (c) comprises step:

(c1) it is relevant with motion between the second sampling group to detect the first sampling group, this first sampling group is comprising locational first sampling identical with current interpolation sampling, sampling before this first sampling and the sampling after this first sampling and in previous row and back delegation with the previous field of the first sampling par locational sampling in, this second sampling group is comprising that identical with current interpolation sampling locational second takes a sample, second the sampling before sampling and second the sampling after sampling and previous row and the back delegation in second the sampling the locational sampling of par next in;

(c2) work as the previous row of front court and the difference between the sampling in the next line comes detection of vertical relevant by calculating in that par is locational with current interpolation sampling;

(c3) by calculate first sampling and second sampling and with par locational when the front court previous row and the sampling in the next line and between difference come detection time vertical relevant.

Images