JP2002510946A - Converter to improve HDTV (High Definition Television) - Google Patents

Abstract

(57) Abstract A standard video signal having 59.94 fields per second is added to a plurality of video fields in a sequence of 1000 video fields into an HDTV video signal having 60.00 fields per second. Electronic device to switch. In the human eye, the device detects the best moment to add a new video field and detects the best image transition motion conditions that occur when the image transition is high or very low. Avoid sudden changes in the video image. To add a new video field, the device uses interpolation techniques to generate two interpolated fields instead of one detected existing video field. The device consists of a de-interlacing module for de-interlacing 60 Hz video images by using state-of-the-art interpolation techniques to calculate missing scan lines. The proposed technique involves directional interpolation of missing scan line pixels in various directions, and also the selection of the best interpolation direction to generate each of the missing scan line pixels. The corresponding de-interlacing device performs the above interpolation in all interpolation directions, and then, based on the quality of the performed interpolation, determines the best direction to perform the interpolation for each interpolated pixel. Consists of state-of-the-art direction detectors to choose from.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

【Technical field】

The present invention relates to video line interpolation, field rate conversion, and a non-interlacing method and apparatus for converting an interlaced standard video image to a de-interlaced video image.

[0002]

[Description of Prior Art]

A quality enhancing converter is an interesting device for broadcasting interlaced standard HDTV resolutions or formats into HDTV video signals. The HDTV quality enhancement converter is essentially divided into three main parts:
It can be decomposed into a field rate converter, a de-interlacing device, and a picture resizing device. The combination of the three modules is 16
Provides a video signal at a field (interlaced) or frame (progressive) rate of 60.00 Hz output with an aspect ratio of: 9.
The main component of the HDTV quality enhancement converter is a de-interlacing device that converts the incoming video signal to a progressive video output. With the progressive output, vertical resizing can now be easily performed. The present invention relates to H
Since the focus is on the first two modules of the DTV quality enhancement converter,
Provided are improved methods and devices for field rate conversion and de-interlaced video images.

In today's broadcast industry, more and more videos, such as movies and documentaries, are being broadcast in the original interlaced video format with a 59.94 field rate per second or typically 60.00 fields per second. And a different image quality aspect ratio of 1 instead of the 4: 3 aspect ratio for the standard format.
The need to convert to a new HDTV video format with 6: 9 has been sought, and a number of solutions have been proposed to do the conversion, all of which provide only partial solutions to this problem. I haven't.

To increase the frequency of a field of a video image from 59.94 Hz to 60.00 Hz, one new field must be added into each of the 1000 sequences of existing fields. Said action often results in images that are still visible to the human eye for a longer period of time, so that reliable results are not obtained1.
This is done by repeating the video fields. The second problem, de-interlacing video images, is solved by inserting new video lines between existing interlaced video lines.

In US Pat. No. 4,677,483 and British Patent No. 2197152A, for interpolation,
Fixed air / temp filters are recommended, but the conversion technique introduces artifacts such as line flicker and diagonal-contour boundary (hereinafter edge) jagged teeth. . In U.S. Pat. No. 4,636,857, a vertical filter and a temporal filter are used to match the image transition motion detector output. US Patent 49890
In 90, an application is made by combining the filtering results of a vertical linear filter and a temporal median filter in an appropriate manner. Disadvantages are the resolution loss of the moving part of the image and the staircase effect on the diagonal edge that moves. US Patent 5
In 001563, the interpolator is an air-phase / time-phase midline filter,
Within the filter, spatial inputs are selected along the direction of the edge being estimated. Even with the techniques described above, there are some disadvantages, i.e. certain edge detectors are not reliable for noisy or high frequency signals. This phenomenon creates temporal flicker because the application at a given edge may be different from frame to frame. The recommended spatial interpolation still causes some resolution loss in parts of the image. A midline filter that is good at preserving edges can produce unnatural results. US Pat. No. 5,019,903 and US Pat.
Within 347599, interpolation is edge-based and purely spatial, producing similar failures mainly due to edge detection techniques. Further, when the number of edge directions considered is small, for example, 45 degrees, 90 degrees, and 135 degrees
In the degree (3) direction, the staircase effect becomes noticeable for nearly horizontal and long edges in the image.

The structure of the de-interlacing device of the present invention is similar to that disclosed in commonly assigned US patent application Ser. No. 08 / 916,960 and Canadian patent application No. 2185061, but with substantially horizontal edges. New edge direction detection techniques are contemplated herein.

[0007] In the case of a 59.94 to 60.00 Hz field rate converter, the references cited are very limited. At the end of the 1000 incoming fields, when a slow transition or still image is detected within the time frame, a frame repetition technique is used. In general, frame repetition can create noticeable motion discontinuities.

[0008]

DISCLOSURE OF THE INVENTION

Accordingly, it is an object of the present invention to provide a converter device for converting an interlaced video signal to a progressive HDTV video signal.
In particular, the present invention provides a method for de-interlacing a lower frequency enhanced video frame frequency of an interlaced video signal later by a second device of the present invention, the de-interlacing apparatus. A converter for converting the intermediate video signal to a higher frequency.

According to a preferred embodiment of the present invention, the apparatus comprises a scan line interpolation means for generating at least one of the video input signals which are temporally and spatially interpolated. The interpolated signals are: a) still part of the image, b) moving horizontal edges, and c
) Suitable for various directions along the edge. The device comprises various means for detecting the condition. In particular, the means for detecting the direction of the edge is robust to nearly horizontal edges in the presence of noise or high-frequency image parts. The proposed edge direction detection method consists of two steps: a) interpolation in all possible directions, and b) selection of the "best" direction between the interpolation results. The possible directions are chosen according to the knowledge of the human visual system.
The device also converts progressive images with 4: 3 aspect ratio into 16: 9 HD.
Comprising means for converting to TV format and resizing, said method comprising:
Since it is a known technique, it is not shown in the following text.

In accordance with another broad aspect of the present invention, there is provided an apparatus for generating a 60.00 Hz field (frame) rate HDTV format video signal from a 59.94 Hz interlaced video input signal, for example. Is done. The apparatus comprises image detection and insertion means for detecting an appropriately smaller number of incoming fields and then inserting a larger number of new frames to produce an accurate 60.00 Hz frame rate. The two new images are preferably interpolated from the incoming signal, rather than being repeated frames.
The proposed "one detect and two insert" technique can reduce discontinuities in image transitions. The device also comprises various detection device means for subjectively determining advantageous conditions in detecting / inserting an image. 50.0
In a 0 Hz system, no image frame rate converter is required.

It is an object of one aspect of the present invention to de-interlace video in which diagonal or near-horizontal, transitional or fixed edge blurring can be reduced. It is to provide a device.

It is an object of one aspect of the present invention that scan line interpolation and the direction of edges that are nearly horizontal or not are robust in the presence of noisy or high frequency image portions. It is to provide an apparatus and a method for scan line interpolation.

It is an object of one aspect of the present invention to provide an apparatus and method for scan line interpolation in which the high resolution in a still image portion is completely preserved.

It is another object of one aspect of the present invention to provide a scan line interpolation apparatus and method that can reduce flicker on horizontal edges that transition vertically.

It is another object of one aspect of the present invention to provide an efficient 59.94 to 60.00 Hz field rate converter with minimal transitional motion discontinuities. That is.

According to a broad aspect of the invention, there is provided an apparatus for generating a signal of an HDTV format video component from an input of an interlaced video component signal.

For a 60 Hz system, the device consists of a 59.94 Hz to 60.00 Hz field rate converter and also a resize device. For a 50 Hz system, the device comprises a scan line overlap device and also a resize device.

The proposed converter from 59.94 to 60.00 Hz consists of two main features: a) detection of frame insertion conditions, and b) frame interpolation and frame insertion.

Detecting the insertion condition in the time frame of the 1000 incoming frames is the first case that occurs: a) a nearly still image, and b) scene changes. , C) an image of the decelerated transition activity, and d) the end of the time frame. The first three conditions are based on the proposed measurement of the locomotor activity index, which is simply the average of the absolute frame difference values.

The proposed insertion of a frame is performed by replacing two newly interpolated frames with one incoming frame. The two frames simply result from a properly separable vertical temporal linear interpolation.

The technique of line doubling is a combination of three main interpolations: a) between the time complements of the stationary parts of the image, and b) vertical transition motion. Vertical interpolation for horizontal edges, and c) between air / time complements directed along the edges. The third proposed interpolation can be further divided into two categories,
A) purely directional along well detected edges and b) directional perpendicular to weakly detected edges. The last type of interpolation is a compromise for near vertical weak edges.

Relevant detection devices for interpolation are: a) detection of a transition using a four-field transition, b) detection of a vertical transition relative to a horizontal edge, and c) edge detection. The proposed edge detection is extended to the next nine edge angles, i.e. 90, 45, 30, 30 and 7, suitable for the human visual system.
Degrees, -4 degrees, -7 degrees, -30 degrees, and -45 degrees. In addition, edge detection is performed in two steps: a) interpolating the image in all predetermined directions, and b) selecting a direction within the variability of the arbitration results. It is. Various skew calculation schemes have been proposed to obtain the most robust decision, even in the presence of noise or high frequency image parts. Judgment as a compromise for weak and nearly horizontal edges has also been proposed.

In a preferred embodiment of the present invention relating to a field frequency converter, the incoming interlaced video signal is converted from a lower field frequency such as 59.94 Hz to the 60.00 Hz required for the HDTV video format. A method and apparatus for converting to a higher frequency such as are provided. The method and apparatus is implemented by adding an auxiliary video field at each sequence of a predetermined number of fields, such as 1000 fields, to increase the frequency of the fields. In fact, as mentioned above, two fields (one frame) can be detected from the sequence of fields, but the four other interpolated video fields added to the sequence of fields are Since it plays the role of generating, the number of fields is increased by two fields (one frame). The process raises the frequency of the field from 59.984 Hz to 60.00 Hz. The same technique can be applied to progressive video signals, in which case the device will use frames instead of fields to be deleted or added.

The preferred embodiment of the invention also relates to the exact moment when the field (or frame) insertion process is performed. The best moment to carry out the process in order for the human eye to be unaware of sudden changes is when the video image is stationary or moving very quickly. The transitional motion detector detects a transitional motion index that is continuously processed to detect the best moment for adding a field to the image. The counter detector can track the number of fields and command the insertion of additional fields for each of the 1000 fields, even if the best insertion condition did not occur until the 1000 fields were reached. , The frequency shift is performed continuously for every 1000 field sequence.

In accordance with the present invention, there is provided a video frame frequency converter for converting a standard video signal having a first frame frequency into one intermediate video signal having a higher frequency, wherein a frequency of the video frame is provided. A converter input for receiving the standard video signal; analyzing the standard video signal; and converting at least one frame in the sequence of existing frames of the standard video signal. Frame insertion detector means for detecting the best moment to add; and means for generating an accelerated video signal from the standard video signal, wherein the accelerated video signal is higher than the higher video signal. Having at least one frame having a second frequency and at least one frame having a predetermined period. Insertion means for inserting into the sequence of existing frames of the accelerated video signal to increase the number of frames of the sequence, wherein the insertion means has a higher frame frequency. And a converter output for providing the intermediate video signal.

Another object of the invention is to provide a standard video signal having a first frame frequency,
Providing a method for converting to an intermediate video signal having a higher second frame frequency, the method comprising: accelerating the standard video signal from the lower frame frequency to the higher frame frequency. Generating an accelerated video signal; analyzing the standard video signal to detect the best moment to add at least one frame into the existing sequence of frames; Generating a control signal, and, upon receiving control of the insertion signal control signal, immediately adding at least one frame into the sequence of existing frames to generate an intermediate video signal. Become.

Still another object of the present invention is to convert a standard interlaced video signal having a first frame frequency to an intermediate interlaced video signal having a higher second frame frequency. Providing a video signal frame frequency converter, the video frame frequency converter comprising: a converter input for receiving the standard interlaced video signal; and Means for generating an accelerated interlaced video signal, the accelerated video signal having the higher second frame frequency, and existing adjacent fields of the standard video signal. By interpolation, at least two new interpolated A frame interpolator means for generating a field; and inserting said two newly interpolated fields into said sequence after inserting an existing frame into said second interpolated video signal. And inserting means for outputting the sequence of frames consisting of the interpolated fields at the higher frequency as the frequency.

According to a preferred embodiment of the present invention, at least a predefined set of directions from which to generate a spatial direction control signal used to perform phase-to-time complementation in the best direction. An improved edge direction detection detection device to be used in a video non-interlacing device for detecting one of the best directions is provided, wherein the edge detection device comprises an input of the detection device. A directional interpolator for performing an interpolation of each one of the preset directions using the past and future video field signals received at the location, for the pixels, the Directional interpolator means for outputting an interpolated signal comprising a signal interpolated for a direction, and interpolating using the interpolated signal received from the directional interpolator means. Consisting edge direction selector means for said at least selected one of the best directions for.

In accordance with a preferred embodiment of the present invention, there is also provided an image quality enhancing converter device for converting a standard interlaced video signal to a progressive HDTV video signal, said image quality enhancing converter comprising: A video frame frequency converter for converting said standard interlaced video signal having a frequency to an intermediate interlaced frequency having a higher frame frequency, said video frame frequency converter comprising: Comprising a converter input for receiving the standard interlaced video signal, and means for generating an accelerated interlaced video signal from the standard interlaced video signal. And the Generating at least two new interpolated fields by interpolating the video signal having the higher second frame frequency and existing adjacent fields of the standard interlaced video signal. Frame interpolator means;
New interpolated fields are inserted into the sequence of existing frames to produce the intermediate interlaced video signal as the higher second
Insertion means for outputting, at a frequency, the sequence of frames containing the interpolated fields, and also analyzing the standard video signal to determine at least one new one in the existing sequence of the standard video signal. Frame insertion detecting means for detecting a best moment for adding a frame, wherein said frame detecting means generates an insertion frame control signal when said best moment is detected. And from a set of predefined directions, at least one best directional control signal to generate a spatial directional control signal that is used to perform space-time complementation in said best direction. To detect direction, it should be used in a video de-interlacing device which receives said intermediate interlaced video signal at the input. An edge direction detection device, wherein the edge detection device is configured to provide, at an input of the device, one of each of the predefined directions using past, present, and future video field signals. Directional interpolator means for performing interpolation on the pixels, said directional interpolator means comprising, for each pixel, an interpolated signal comprising an interpolated signal in said predefined direction. And an edge direction selector means for selecting said at least one best direction for interpolation using said interpolated signal received from said direction interpolator means.

The invention will be better understood from the following description made with reference to the drawings, in which:

In the drawings, FIG. 1 a shows three main parts of the quality enhancing converter 10. The first part of the system is a video frame frequency or field rate converter 12 which, according to a preferred embodiment of the present invention, has a lower frequency, such as a 59.94 Hz interlaced video signal input 14. And then outputs a higher video frame frequency signal, such as a 60 Hz interlaced video signal 16, said video signal 16 having exactly 60 video fields per second. So
Meets HDTV field frequency standards. The intermediate video signal 16 is then:
The deinterlacing device 20 has the function of providing a progressive video signal 22 having the same input image aspect ratio of 4: 3 as the signal 14. With the progressive video signal 22, the resizing device 24 can easily read the input signal 2
2 to a progressive HDTV video signal 26 having an aspect ratio of 16: 9.
Can be converted to The resizing device 24 is mainly composed of separable vertical and horizontal digital interpolation filters. Because filter technology is well known, the resizing device 24 will not be described in detail herein.

[0032] The progressive HDTV video output 26 can be converted to an interlaced HDTV video by deleting appropriate scan lines in each screen or image as necessary.
The signal 28 can be further converted. The selective deletion of the scan line is shown in FIG.

The present invention applies to the European version of the HDTV standard. In the standard, HD
The TV signal has a frequency of 50 Hz. Because the reference interlaced video signal 30 has the same frequency, the field frequency converter 12 is no longer needed in the system. As shown in FIG. 1 b, the interlaced video signal 30 is provided directly to the de-interlacing device 20. In other words, according to the preferred embodiment of the present invention, de-interlacing device 20 functions for all field rate video inputs.

FIG. 2 shows a block diagram of a proposed de-interlacing device for interlaced digital luminance video signals. The present invention provides an improved edge direction detector 44 that is part of the overall de-interlacing device 20. The improved system relies on pure vertical interpolation for vertically transitioning edges, and for the general case, an edge-oriented airspace It is an adaptive interpolator that combines with time complements. Further, as shown in FIG. 3, the number of edge directions is up to 9, that is, 4 degrees, 7 degrees, 30 degrees, 45 degrees, and 9 degrees in the plus direction.
In the negative direction, it is extended to -4 degrees, -7 degrees, -30 degrees, and -45 degrees. The proposed directions have been selected in an almost logarithmic order according to the human visual system. In fact, the image results are more comfortable if near horizontal edges are carefully interpolated.

Returning to FIG. 2, the interlaced video input 005 is applied to two serially connected delay devices 32 and 33. The two field delays provide two delayed video signals 34 and 35, respectively. According to a preferred embodiment of the present invention, the video signals 35, 34 and 30 indicate past, present and future video fields. The signals are transmitted to the interpolators 38, 40, and 42 and the three detectors 44, 46, and 48 in a suitable manner. The cited detector, in turn, controls the qualification of the system to provide the final interpolated signal output.

The first interpolator 38, the time interpolator 38, provides an average signal from the past frame signal 35 and also from the future frame signal 30. More precisely, the temporal output 50 is given by the following equation.

(Equation 3) In the above equation, B is the next field pixel value spatially corresponding to the pixel to be interpolated, and C is the pixel value at the same position in the previous field. The locations of the various pixels for the interpolation are better illustrated in FIG.

The vertical interpolator 40 accepts an existing field video signal 34 only when it is input. The output 40 of the vertical interpolator specified by VF is also given by:

(Equation 4) In the above equation, A and A ′ are respectively adjacent pixels corresponding to the previous and next existing scan lines in the horizontal direction of the pixel to be interpolated, and F is adjacent to the pixel A in the vertical direction. The number of existing pixels. F is adjacent to A vertically as shown in FIG.

[0038] Directed air / time interpolator 42 receives the three interlaced video signals as inputs, ie, field signal 34, past field signal 30, and future signals. It is accepted as a field signal 35 and, at the same time, as a control signal 54 output by the edge direction detector 44. Depending on the state of the control signal 54, the air / time complement device 42 to which the direction is directed calculates one of the following thirteen equations.

(Equation 5)

The first four equations, ie, Equations 3 through 6, are calculated when the condition “Mix” in the control signal 54 is off or zero. The last four equations, ie, Equations 12 to 15, are calculated when the condition “Mix” is ON or 1. Finally, the other equations, ie equations 7 to 11, are not dependent on the condition "Mix". The first nine expressions,
That is, A ± i, A ′ ± i, B ± i, B ′ ± i, C ± i, C ′ ± i, D ± i, D ′ ±
i, E ± i, E ′ ± i, F ± i, F ′ ± i, G ± i, G ′ ± i, H ± i, and H ′
See FIG. 4 for a better understanding of the location of the pixels used in the calculations within ± i.

Equation 1 is chosen to reduce the additional noise by a factor of 3 dB. Equation 2 shows the simplest four-tap half-band filter, and Equation 3
To 10 are edge directed versions of the vertical temporal half band filters defined by the following equation: That is,

(Equation 6)

Applicants have discovered that the air-phase filter provides more comfortable results than those obtained with a unitary air-phase filter.

Here, referring to Equation 11, SST 90 representing vertical interpolation is also a vertical temporal filter similar to that defined in Equation 16, but the vertical phase filter in Equation 11 The frequency band is greater than that defined in Equation 16 when the temporal frequency is approximately zero. Said features were selected because the human visual system is more sensitive to the stationary parts of the image. Moreover, the filter described in Equation 11 is disclosed in US patent application Ser. No. 08 / 916,960 mainly because the filter has better vertical band and sharper transitions in and out. Is different from the VT filter used in

When the detection result for an almost horizontal edge is found not to be sufficiently reliable,
Interpolation of equations 12 through 15 with the "Mix" condition "on" has been proposed as a compromise technique.

Referring again to FIG. 2, two interpolated signals VT and SST, indicated by reference numerals 56 and 58, respectively, are controlled by a binary signal 62 provided from a vertical movement detector 46. Is supplied to the selected selector 60. When the control binary signal 62 is “ON”, the selector 60 outputs the vertical interpolator VF output 5
The SF signal 64 selected as 6 is output. Otherwise, when the binary signal 62 is "off," the multiplexer 60 selects the air / time complement device output 58 to be directed.

The selector SF output 65 and the output 50 of the interpolator TF are combined in a temporal adapter 66 to produce the final interpolated video signal for the nonexistent scan line of the interlaced output signal. 68 is provided. The time phase adapter 66 is controlled by a transitional movement instruction value 70 sent by the time phase transitional movement detection 48.

Finally, the interpolated video scan line signal 68 and the existing video scan line signal 72 are combined by a multiplexer 74 to generate a progressive luminance signal 22.

The detection device 44 for edge detection, the vertical transition motion detection device 46, and the temporal transition motion detection device 48 work in conjunction with the interpolation technique described above.

The purpose of the temporal phase movement detector 48 is to find rapidly moving or nearly stationary parts of the image. To perform the locating operation, the device uses as an output the low-pass filtered video signals 76 and 77 instead of the original 30 and 35 signals, which may be noise signals. .

The purpose of the vertical transition detector 46 is to find a moving scan line in a video image sequence.

The edge direction detector 44 has two functions: first, it selects the best direction to perform directed interpolation among the nine possible directions. The second function is to compute a compromise interpolation for detected substantially unreliable near horizontal edges. The decision process is performed in two stages: the first, the image is interpolated among all possible directions, and the direction with the least variation is selected. Nine directional interpolators 80 receive as input three video signals 30, 34 and 35 from past, present and future fields. The interpolation is described by equations 3, 4, 5, 6, 7, 8, 9, 9, 10, and 11, each of which has nine possible directions, -4. Degrees, 4 degrees, -7 degrees, 7 degrees, -30 degrees, 30 degrees, -45 degrees, 45 degrees, and 90 degrees. The resulting nine interpolated output signals 82 [a-i] are applied to an edge direction calculator 84 and to a horizontal and vertical high frequency detector 86.

FIGS. 5 and 6 show directions 90 degrees, 45 degrees, 30 degrees, 7 degrees, 4 degrees, −45 degrees, and −3 degrees.
FIG. 4 shows a detailed diagram of edge direction calculation for detection of 0 degrees, −7 degrees, and −4 degrees. In the figures, each interpolated input signal, such as 82a, is sent to a horizontal low-pass filter 90 to eventually remove non-interlaced enhanced horizontal edges. The impulse response of the linear phase filter 90 is (1, 3
, 4, 3, 1). Filter outputs 92, 94, 96, 98, 100;
Also, 102, 104, 106 and 108 may be individually referenced 11
Multiplied by the direction variation calculator, which is scaled from 0 to 126. In general, each calculator is a directional high-pass filter whose impulse response is described in FIGS. The computer outputs 110 to 126 are denoted by reference numerals 130 to 14
6 is sent to the waved magnitude device to convert the initial value to the magnitude of the variability in the nine possible directions. The output of the absolute value device is sent to low pass filters 151-166 associated with the device to smooth out any noise that may occur. Further, in order to favor the vertical direction (90 degrees), the low-pass filter may be only a vertical filter having a gain that is generally twice as low as one of the other directions. The low-pass filters for the other directions can be the same, consisting of vertical and horizontal filters.
The impulse responses of the filters are illustrated in more detail in FIGS.

Referring again to FIG. 2, the outputs 82 [ai] of the direction interpolator 80 are sent to the horizontal and vertical high frequency detector 86 as described above. The purpose of the high frequency detection device is to find high horizontal and vertical image data areas where errors can be introduced in the process of estimating the edge direction. Thus, it may be necessary to change the detection device 86 for possible directions.

FIG. 7 shows a high frequency detection device 86 a for the vertical direction according to a preferred embodiment of the present invention. The corresponding interpolated signal 82a is detected by a detection device 8 which detects the intensity of the high frequency by applying a Laplacian 170 and then an absolute value device 172.
6 is multiplied. The output 174 of the device indicates the magnitude of the high frequency signal,
It is applied to a detector 176, which is simply a level comparator. The comparator output is a binary signal 178, which is equal to 1 if the input signal 174 is greater than a threshold 180, otherwise the binary signal 17
8 is equal to 0. The threshold value 180 according to a preferred embodiment of the present invention is set to 40. The binary output signal 178 is a moving window 3X3.
Are sent to a skew computing device that can link the isolated detections within each other. The details of the described skew operation device are described in the same FIG. 7, i.e., the device consists of a delay 182, a filter 184, a comparator 186, and an OR gate 188. be able to. The gate output signal 190a represents a binary map of the high frequency region for a direction of 90 degrees.

FIG. 8 shows a high frequency detection device 86 for directions of 45 degrees and −45 degrees. In the detectors according to a preferred embodiment of the present invention, the pixels are input signals 82b or 82f except for a specific high frequency pattern corresponding to a possible direction.
If the horizontal or vertical frequency component in the data exceeds a certain threshold, it is written that the frequency is in the high frequency range. For the 45 degree direction, the horizontal and vertical high frequency components are:
They are detected by high-pass filters 192 and 194, respectively. On the other hand, the mask 196
Defines a specific pattern for the 45 degree direction. For the -45 degree direction, high pass filters 198 and 200 perform the same function while mask 202
Defines a -45 degree pattern. The impulse response of the filters and, at the same time, the details of the detection process are detailed in FIG. The output of the binary signal of the detector is 190b and 1 for directions 45 degrees and -45 degrees, respectively.
90f signal.

FIG. 9 shows the high-frequency detection device 86 for the directions of 30 degrees and −30 degrees. 2
The detection devices 86 are the same as those shown in FIG. The only difference is
It is a specific high frequency pattern. For detection in the direction of 30 degrees, the detection is:
While performed by the filter 208, the angle -30 degrees is performed by the filter 210. The output of the binary signal of the detector is 30 degrees in the direction and -3 degrees.
For 0 degrees, they are 190c and 190g, respectively.

FIGS. 10 and 11 show four similar detectors 86 for four substantially horizontal directions of 7 degrees, -7 degrees, 4 degrees, and -4 degrees. As an example, for a direction of 7 degrees, the horizontal filter 212 detects a horizontal high-frequency component. The vertical filter 214 does the same job for vertical high frequency components. Further, a vertical mask 216 can be used for a particular pattern. The detection device 86
The signals 190d and 19d for degrees, -7 degrees, 4 degrees, and -4 degrees, respectively.
0h, 190e, and 190i are output.

Referring again to FIG. 2, the output 220 [ai] of the edge direction calculator 84 and the output 190 [ai] of the high-frequency detector are shown in FIG. Sent together to direction selector 222. First, nine edge direction variations 220 are applied to a minimum selector 224 that outputs data associated with the minimum value, a first signal 226 comprising the corresponding direction, and a second signal 228 having the corresponding data. Sent. If two or more equal minimums are detected, the device 224 will determine 90 degrees, 45 degrees, -45 degrees, 30 degrees, -30 degrees, 7 degrees, -7 degrees, 4 degrees, and- With four priorities, only one direction is selected.

The first signal 226, the second signal 228, and the 90-degree variation signal 220a
Pseudocode is applied to the logical device 230, better shown in FIG. The logic device 230 selects only the minimum (first minimum) direction or the vertical direction. The second minimum is used for matching comparison purposes. The logic device 230 has 90 degrees, 45 degrees, 30 degrees, 7 degrees, 4 degrees, -45 degrees, -30 degrees, -7 degrees, and -4 degrees, respectively. 9 binary outputs 232
a and 234 to 248 are provided. The output in the selected direction is set to "1" while the others are set to "0". Finally, the possible pixels are
The selected direction can be reset to "0" if it is in a frequency range where the reliability detected due to the presence of the high frequency binary signal 190 cannot be placed. Except in the vertical direction, each of the eight selector outputs therefore performs a negation of the associated detected high-frequency binary signal and a different AN via 250h.
This is made effective by the D gate 250a. The nine resulting directions, each of the possible directions, 90 degrees, 45 degrees, 30 degrees, 7 degrees, 4 degrees, -45 degrees,-
Outputs 232a through 232i indicate one of 30 degrees, -7 degrees, and -4 degrees.

In general, the edge direction selector output signal 232 [a-i] is pointed and consists of a number of irregularly discrete directions or discontinuities along major edges. Therefore, it is necessary to enhance the detection result before making a final decision. The output 232 [a-i] is, as shown in FIG.
It is transmitted in the filter transmission 127, and performs a judgment operation to make a decision to be made. FIG.
As best shown in 4 and 15, except for the vertical direction (90 degrees), the binary signal 232 [bi] can be skewed in four or five consecutive steps. . The five steps can be well described as horizontal, direction, vertical and another horizontal, and finally a logical vertical skew operation. The last operation is used only for the substantially horizontal direction, ie, 7 degrees, -7 degrees, 4 degrees, and -4 degrees. For the first four stages, each one can consist of eight binary filters followed by a level determining device acting in parallel. Each filter and its associated decision device can only be used for one direction. The filter mask and level detector are described in FIGS. For example, the first skewer calculation filter 2
The impulse response of 60 is (1,1,1,1,1), in which the central coefficient corresponds to the position of an existing pixel. Detector 264
Is set to 1 = 2. The skewering operation 2 shown in FIG. 15 has the same structure as the first skewering operation. The skew operation 5 used for the substantially horizontal direction consists of four logic devices for four possible directions. Each filter is a linear horizontal filter, followed by a logic device that makes use of the code described in the pseudo code described in FIG. 14 to provide two binary outputs 266 and 268. As described above, "
The output 268 [d, e, h, i] corresponding to the direction, called Mix ", can be indicated by some compound conditions in the output image scan line interpolation.

Here, FIG. 18 showing the final direction determination 270 in the pseudo code will be referred to.

The input signal for the determination is made up of two input vectors 135 [bi] and 150 [d, e, h, i] for the direction and the mixed condition, respectively. The final decision output 54 indicates the selected interpolation with or without mixed conditions,
As shown in FIG. 2, both are transmitted to the oriented phase / time complement device 42 and to the vertical transition motion determination device 272. For the mixed condition, the interpolator 42 combines the direction and the vertical interpolation into the average as described in Equations 12-15. Otherwise, the interpolation is strictly directional interpolation. Further, within the vertical transition motion device 272, only the directional information transmitted by the signal 54 can be considered.

As shown in FIGS. 20 and 21, respectively, the vertical movement detector 46 and the phase detector 48 remain unchanged and have already been enclosed herein by reference by the same applicant. Part of the preferred embodiment of the present invention relating to the device described above is complete as described in US patent application Ser. No. 08 / 916,960.

FIG. 21 illustrates an adaptive scan line overlap technique for video image luminance components. Applicants have found that adaptations based on moving or stationary parts of the image,
It has been found sufficient for nearly horizontal edges. The structure remains unchanged compared to the one proposed in the aforementioned US patent. The only difference is
11 is a fixed TV interpolation filter 280 described in Equation 11.

FIG. 22 shows a frame frequency converter or a field rate converter 1
2 shows a preferred embodiment of the invention with respect to FIG. Even if the following paragraph
The same technique is also used to convert the frame frequency of a progressive video signal from a lower frame frequency to a higher frame frequency, even though it primarily describes a frame frequency converter for a standard interpolated video signal. You have to keep in mind that you can. Said special features, although rarely needed, still serve to convert the reference progressive video signal to HDTV progressive video with a higher frame frequency.

In a preferred embodiment of the present invention, the frame frequency converter 12
. A reference interpolated video signal 14 having a frequency of 9400599402 fields is received. The standard video signal 14 is simultaneously transmitted to a clock and field synchronization generator, a buffer memory 302, and a frame counter 304.
, And also supplied to the frame insertion detection device 2107.

The buffer memory 302 is the means used to generate the accelerated video signal 308, the 59.94 Hz field synchronization control signal 30
6, the digital video input signal 14 is read, and then at the output of the memory, a 60 Hz field synchronization control signal 307 is used.
Sending the accelerated video signal 308 having a field frequency of
It can be a (first in first out) device. An accelerating means, called buffer memory 302, can also be used to freeze the video input frame and accept an insertion frame control signal 310 to balance the output video speed. The clock and field synchronization generator 300 controls the control signals 306 and 30
7, while the frame insertion detector 312 sends out the control signal 310.

The frame counter 304 receives 0 to 1 after receiving the video signal input 14.
Count 000 incoming frames. The counter provides a time frame or window in which one new video image is inserted to reach a total of 1001 images for each incoming sequence of 1000 images. The 1001/1000 ratio is 60H from the initial 59.94Hz frequency.
Required to provide a video output of z.

The functional block diagram shown in FIG. 22 shows the proposed frame insertion detection device 15 for the case of an interpolated video signal. The purpose of the detection device is to determine the correct moment to insert a frame in a predetermined sequence of 1000 consecutive existing frames. Said frame,
Although it is possible to copy from an existing frame, it is preferable that the frame be newly interpolated from an adjacent existing frame if possible. In order to reduce the prominent artifacts that can be created when inserting a new image, the proposed detector 15 can have the sequence of frames examined to detect the following situations: A) a sequence of static or near-static images; b) sudden scene changes; c) slowed down movement activity; and d) the end of the time frame. The said static situation a) allows the newly generated image to be easily interpolated,
it is obvious. The scene change b) is again understandable, since in this case the artifact of the image interpolation is not noticeable to the human visual system. Situation c) is a compromise technique for dynamic image sequences, i.e., insertion occurs when transitional motor activity is reduced to an adaptively varying threshold. The situation d) is self-evident and occurs when no other situation has occurred until the counter reaches 1000.

The standard interlaced video signal 14 is provided to a transitional motion index calculator 320, shown in FIG. 24, which evaluates the absolute difference between frames. The value, referred to herein as the transitional motion index 322, is provided to a fixed threshold determination device 324, a scene change detector 326, and an adaptive threshold determiner 328. The fixed threshold determination device 324 is simply a level detection device, better shown in FIG. 26, which device is used for static or nearly static image detection. The device provides an FT output 330 which can be one of the signals that a possible image insertion condition has been reached. The scene change detection device 326 also provides a binary output 332 when the difference in the transition kinematic index between two consecutive frames is greater than or equal to a predetermined threshold. A block diagram of the corresponding scene change detection device is shown in more detail in FIG. The adaptive threshold 328, shown in more detail in FIG. 26, may comprise a first order low-pass filter 334 providing an output 336. The modified weighted value 338 with the possible coefficient 0.9 coming from the low-pass filter 336 is used as an adaptive change threshold for the detector 340. If the transitional motor activity index of the instantaneous frame is less than the threshold signal 338, the output of the detection device AT is "on" providing a signal for a possible image insertion instant.

The function of the multiplexer 334 is to quickly change the changing threshold when a sudden scene change is detected. Returning to FIG. 23, the frame count detector 350 provides a binary signal 352 consisting of count = 1000 signals for image insertion time limitation. The frame count detector 350
Also provides a count = 0 signal 354 for the beginning of the time interval. The signal is sent to the logic device 356 while the four binary signals 332
, 330, 342, and 352 are provided to an OR gate 358 to provide an output signal 360 that specifies a possible image insertion signal. The signal 360 is also transmitted to the logic device 356 to provide an insertion frame control signal 310 having an "on" value only once during a time window of 1000 frames.
In fact, the frame insertion detector 15 implements the first of the following four possibilities: FT, SC, AT, and count = 1000. The frame control signal 310 is transferred to the buffer memory 302 for freezing conditions and to the multiplexer 364 to select an interpolated image 336 from the frame interpolator 368.

FIG. 27 illustrates the technique for the proposed frame interpolator module 368 and for the insertion means or frame insertion module 364. To reduce the discontinuity artifacts in the transitional motion that occur when a new, interpolated image is introduced, the incoming single frame is detected when the insertion condition is detected,
The proposed frame insertion technique by replacing two new interpolated frames (if interlaced, replacing two incoming fields with new four interpolated frames) Is performed according to a preferred embodiment of the present invention. FIG. 27 also shows various field positions that are continuous in time. The existing frame 400 consisting of two frames C and D will be deleted. Two new frames will be inserted, a first frame consisting of one of fields P and Q and a second frame consisting of one of fields R and S. The distance in time between the previous field B and the interpolated field P is set to 3/5 of the interval of the incoming field according to the best mode of the invention.
It can be. The normalized distance between P and C is therefore 3/10. To simplify the description, the proposed interpolation technique for field P is:
Only the two closest existing fields B and C are based. Similarly,
The basis must be the fields C and D used for the calculation of the interpolated field Q, and so on.

The new field interpolation can be performed in two separate stages: a
A) vertical interpolation for lost scan lines in existing fields, and b) time complementation for new fields to be inserted. FIG. 28 shows the individual interpolation filters. The reader should note that the pixel and scan line notes are described in FIG. The vertical interpolation for the missing scan line is that the impulse response is provided from the half-band filter described below, ie (-8, 0, 40, 64, 40, 0, -8). . The temporal filter for the new field interpolation is a two-tapped filter, in which the coefficients 3/4 and 1/4 are used and the normalized field distance A practical fixed point value is shown, which is approximately equal to 3/5 and 3/10.
The coefficient 3/4 is linked to the closest existing field to the possible interpolated field. The factor 1/4 is linked to other existing fields used for interpolation. For example, P2 can be described by the following equation.

(Equation 7)

Returning to FIG. 22, the interpolated video image 366 is
0 and can also be selected by multiplexer 364 to provide an output, 60 Hz interpolated video signal 16.

[Brief description of the drawings]

FIG. 1 is a general block diagram showing the main parts of an HDTV quality enhancement converter, in which FIG. 1a shows a preferred embodiment of the invention for a 60.00 Hz system; FIG. 1b also refers to a 50.00 Hz system.

FIG. 2 illustrates a preferred embodiment of the present invention with reference to a de-interlacer.
It is a block diagram which shows a detailed function.

FIG. 3 shows the positions of nominal pixels corresponding to nine possible edge directions.

FIG. 4 shows pixel locations used for interpolation of various scan lines.

FIG. 5 shows the calculation of the direction of the edge for a series of directions corresponding to 90 °, 45 °, 30 °, 7 °, 4 °.

FIG. 6 illustrates the calculation of the direction of the edge for a series of directions corresponding to −45 degrees, −30 degrees, −7 degrees, and −4 degrees.

FIG. 7 illustrates a high frequency detection device for a 90 degree direction according to a preferred embodiment of the present invention.

FIG. 8 illustrates a high frequency detection device for 45 degree and -45 degree directions according to a preferred embodiment of the present invention.

FIG. 9 shows a high frequency detection device for directions of 30 degrees and -30 degrees according to a preferred embodiment of the present invention.

FIG. 10 shows a high-frequency detector for directions of 7 degrees and −7 degrees according to a preferred embodiment of the present invention.

FIG. 11 shows a high frequency detection device for directions of 4 degrees and -4 degrees according to a preferred embodiment of the present invention.

FIG. 12 illustrates an edge direction selector according to a preferred embodiment of the present invention.

FIG. 13 shows an algorithm for minimizing the edge direction selector.

FIG. 14 shows edge binary filters in the skew operations 1, 3, 4, and 5.

FIG. 15 shows directions of 45 degrees, -45 degrees, and 30 degrees. 9 shows an edge binary filter in a skew operation 2 for a direction of −30 degrees.

FIG. 16 shows an edge binary filter in the skew operation 2 for directions of 7 degrees and -7 degrees.

FIG. 17 shows an edge binary filter in the skew calculation 2 for directions of 40 degrees and −400 degrees.

FIG. 18 shows a direction determination block diagram in a pseudo code format.

FIG. 19 is a block diagram of a proposed transitional motion detection device according to a preferred embodiment of the present invention as shown in FIGS. 2 and 21.

FIG. 20 is a block diagram of a proposed vertical transition motion detection device according to a preferred embodiment of the present invention.

FIG. 21 is a block diagram illustrating general functions illustrating a method for a scan line overlap device or chrominance (color difference) component.

FIG. 22 is a block diagram illustrating general performance illustrating a preferred embodiment of the present invention with reference to a field frequency converter.

FIG. 23 is a block diagram showing the performance of the detection device for detecting the frame insertion condition shown in FIG. 22;

FIG. 24 is a block diagram of the transition kinetic index calculator for the embodiment shown in FIG. 22.

FIG. 25 is a high-level flowchart of a scene change detection device according to a preferred embodiment of the present invention as shown in FIG. 22;

26 shows a fixed and adaptive threshold detection device, respectively, according to the preferred embodiment of the present invention shown in FIG. 21.

FIG. 27 illustrates a proposed technique for frame interpolation and frame insertion, also shown in FIG. 21, according to a preferred embodiment of the present invention.

FIG. 28 illustrates a separable vertical temporal filter for frame interpolation according to a preferred embodiment of the present invention.

──────────────────────────────────────────────────続 き Continuation of front page (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE ), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SL, SZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AE, AL, AM, AT, AU, AZ, BA, BB, BG, BR , BY, CA, CH, CN, CU, CZ, DE, DK, EE, ES, FI, GB, GD, GE, GH, GM, HR, HU, ID, IL, IN, IS , JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, UA, UG, US, UZ, VN, YU, ZA, ZW (71) Applicant 2323 Halpern, Saint-Laurent , Quebec H4S 1 S3 CANADA (72) Inventor Poillier, Daniel Canada, Quebec J7V 6Z5, Vaudruille Dorion, Tvet 437 F-term (reference) 5C063 AA11 BA04 BA08 BA09 BA20 CA01 CA05 [Continuation of summary] The above interpolation is performed in all interpolation directions, and then based on the quality of the performed interpolation, it is best to perform the interpolation for each interpolated pixel. Consisting cutting edge direction detecting device for selecting a direction.

Claims (32)

[Claims] 1. A video frame frequency converter for converting a standard video signal having a first frame frequency to an intermediate video signal having a higher second frequency, wherein the converter receives the standard video signal. An input; a frame insertion detector means for analyzing the standard video signal and detecting the best moment to add at least one frame into the sequence of existing frames of the standard video signal. Means for generating an accelerated video signal from the standard video signal, wherein the accelerated video signal has a higher second frame frequency; and at least one frame is preset. Sequence of an existing standard video signal to increase the number of frames in the sequence having Insertion means for inserting the intermediate video signal having a higher frequency, and a converter output for providing the intermediate video signal. A video frame frequency converter. 2. The apparatus of claim 1, further comprising: frame interpolator means for generating at least one existing temporarily adjacent newly interpolated frame, said at least one newly interpolated frame comprising a first frame frequency. The higher second
2. The video frame frequency converter according to claim 1, wherein the video frame frequency converter is used by the insertion means for inserting into the existing sequence of frames to increase the frequency. 3. The interpolation is also performed spatially, and in order to increase the newly interpolated frame, by interpolating vertically adjacent pixels before interpolating temporarily adjacent frames. The video frame frequency converter according to claim 2, wherein the video frame frequency converter comprises: 4. A method for providing a first synchronization signal having a frequency corresponding to the standard video signal and the second synchronization signal, and a second synchronization signal having a frequency corresponding to the intermediate video signal. A clock generation device and a frame synchronization means, for counting frames from one to the number of frames comprised in said sequence of existing frames, and for generating an end-of-sequence signal, Sometimes, the frame counter means for providing to the frame insertion detection device; andthe generating means comprises buffer memory means for storing the existing video signal of the standard video signal, The buffer memory means stores the standard video signal from the input and the buffer memory. Stage, wherein the standard video signal is received directly at a rate corresponding to the first synchronization signal received from the clock generator and the field synchronization means, and wherein the buffer memory means has the higher second frequency. Outputting an accelerated video signal at a buffer memory output according to said second synchronization signal;
3. The video memory of claim 2, wherein said buffer memory means comprises an insert frame control signal received from said frame insert detector means for controlling the output of said buffer memory. Frame frequency converter. 5. The buffer memory means continuously converts the accelerated video signal to the accelerated video signal as soon as the inserted frame control signal is provided by the frame insertion detection device means. Adding a new interpolated frame by either outputting the video signal at a rate corresponding to the second higher frequency or by stopping outputting the accelerated video signal. 5. The video frame frequency of claim 4 wherein the inserted frame control signal causes the buffer memory means to skip outputting a frame when the best moment occurs. converter. 6. The frame inserting means is a multiplexer having two output ports and one input port, and a first input port is output by the buffer memory means according to the second synchronization signal. Receiving the accelerated video signal, the second input receiving the new interpolated frame from the frame interpolating means, and the inserted frame control signal being provided by the frame inserting means. As soon as the multiplexer outputs one of the accelerated video signals received at the first input port and the interpolated frame received at the second input port. Wherein the newly interpolated frame added to the existing sequence of frames increases the frame frequency of the signal. 6. The video frame frequency converter according to claim 5. 7. The buffer memory means stops outputting the existing frame for a predetermined time as soon as the insertion frame control signal is controlled, and at least one existing frame is output. As soon as a frame has been skipped and, at the interval of the period, the inserted frame control signal has been controlled, the multiplexer has interpolated more than the number of new interpolated frames from the frame interpolator means. 7. The video of claim 6 wherein the number of frames is accepted and the new interpolated frame is output in place of the existing frame to add a frame into the sequence of existing frames. -Frame frequency converter. 8. The method of claim 1, wherein the standard video signal is a standard interlaced video signal, and wherein the first frame frequency of the standard interlaced video signal is about 59.94 fields per second. Wherein said intermediate video signal is an interlaced intermediate video signal; and said higher second frame frequency of the interlaced video signal is:
Being approximately 60.00 Hz, and that the sequence of existing frames consists of 1000 video frames, and that to increase the first frame frequency, one newly interpolated frame is 8. The video frame frequency converter according to claim 7, wherein the video frame frequency converter is added to the existing 1000 sequences. 9. The method according to claim 1, wherein the time interval is equal to a period of one frame, and the buffer memory means stops outputting an existing frame during the period of one frame, thereby causing the existing frame to exist. Jumping over one existing frame in the sequence of frames, the multiplexer accepts two new interpolated frames instead of the one existing frame during the time interval. The video frame frequency converter according to claim 8, characterized in that: 10. The method of claim 1, wherein the existing frame comprises two existing video fields, and each of the newly interpolated frames comprises two interpolated video signals. 10. A video frame frequency converter according to claim 9. 11. The two interpolated video signals are pixels calculated by a pixel for each of the pixels interpolated using information consisting of at least two adjacent existing pixels. The video frame frequency converter according to claim 10, characterized in that: 12. A transition kinematics calculator for calculating a transition kinetic index of a video image, the transition kinetic index calculator indicating a level of motion of a sequence of video frames. A scene change detection device for providing a first inserted frame control signal when a change occurs, wherein the abrupt change indicates a scene change, and the scene change detection device directly controls a transitional motion index. And at the input of the threshold device, a slow change or no change is indicated by the transition kinematic index at the input of the threshold device. And a fixed threshold device for providing a third insertion control signal when a drop in sequence transition motion velocity is detected. The adaptive thresholding device calculates an average transitional motion index using a sequence of previous video frames, wherein a transitional motion rate in the thresholding device is constant or substantially constant; and Comparing the average transitional movement index with the transitional movement index received from the transitional movement index calculator;
An inserted frame control signal is output when the average transition kinematics index is lower than the transition kinematics index received from the transition kinematics calculator, and an incoming frame in a sequence of 1000 frames. Calculate
A counter detection device for providing a fourth inserted frame signal when the count of 1000 is reached; and when the first, second, third, and fourth inserted frame signals are output, A logical OR gate for outputting at least one intermediate insertion frame signal; an output from the logical OR gate; and a count = 0 signal from the count detection device, wherein the logical device receives a count = 0 signal. Wherein the count = 0 signal is used to reset the logic device so that the device outputs one inserted frame signal only for every 1000 frame sequence. The video frame frequency converter according to claim 11, characterized in that: 13. A method for converting a standard video signal having a first frequency to an intermediate frequency signal for converting to an intermediate video signal having a higher second frequency, comprising: Accelerating from a lower frequency to the higher frequency to produce an accelerated signal; and detecting a best moment for adding the standard video signal to at least one frame sequence of existing frames. Analyzing to generate a frame insertion control signal, and adding the at least one frame to an existing sequence of frames as soon as the frame insertion control signal is controlled. Generating an intermediate video signal. 14. Computing at least one new frame pixel by interpolation of at least one temporal and empty phase of at least one temporally adjacent frame or an adjacent pixel of an existing frame. 14. The method of video frame frequency conversion of claim 13, further comprising generating an interpolated frame. 15. A method for adding at least one frame, comprising: jumping at least one frame from the accelerated signal as soon as the frame insertion control signal is controlled; As soon as the control signal is controlled, inserting at least two frames instead of the at least one skipped frame to increase the number of frames in the sequence. 14. The method of video frame frequency conversion according to claim 13, wherein: 16. A method for adding at least one frame, comprising: jumping at least one frame from the accelerated signal as soon as the frame insertion control signal is controlled; and As soon as the control signal is controlled, comprising inserting at least two frames instead of the at least one skipped frame to increase the number of frames in the sequence. The method of video frame frequency conversion according to claim 14, wherein: 17. A video frame frequency converter for converting a standard video signal having a first frame frequency to an intermediate video signal having a higher second frequency for receiving an interpolated video signal. Means for generating an accelerated interlaced video signal from the converter input and the standard interlaced video signal, wherein the accelerated video signal has a higher second frame frequency. Frame interpolator means for generating, by interpolation, at least two new interpolated fields according to existing adjacent fields of the standard video signal; and at least two newly interpolated fields. Field is inserted into the existing sequence of frames and interpolated The sequence of the frame consisting of field, a video frame frequency converter, characterized in that it consists of inserting means for outputting at intermediate interlaced video signal similar to a second, higher frequency frequency. 18. A frame for analyzing a standard video signal and for detecting the best moment to add at least one new frame into the sequence of existing frames of the standard video signal. Consisting of an insertion detector,
18. The video frame frequency converter according to claim 17, wherein said frame insertion detection means generates an insertion frame control signal when the best moment is detected. 19. The method according to claim 19, wherein the interpolation is at least one of a temporal phase and an empty phase, and wherein the frame interpolating means generates at least i) a pixel of the field interpolated by a pixel;
Using information gathered from at least temporally adjacent fields of the newly interpolated field; and ii) at least empty of the newly interpolated field to generate pixels of the newly interpolated field by pixels. 19. The video frame frequency converter according to claim 18, comprising using information collected from locally adjacent fields. 20. A method for providing a first synchronization signal having a frequency corresponding to an interlaced standard video signal and a second synchronization signal having a frequency corresponding to the intermediate interlaced video signal. A clock generator and field synchronization means for storing existing video frames of said interlaced standard video signal, said buffer memory means comprising: Receiving the interlaced standard video signal from the clock generator at a rate corresponding to the first synchronization signal received by the buffer memory from the clock generator and also from the field synchronization means; And said buffer memory means comprises a buffer memory means. At an output, an interlaced video signal comprising an accelerated, interlaced standard video signal having the higher frequency received according to the second synchronization signal and received by buffer memory means. 20. The video frame frequency converter according to claim 19, comprising an output. 21. The frame inserting means is a multiplexer having two output ports and one input port, and a first input port is output by the buffer memory means according to the second synchronization signal. Receiving said accelerated video signal, said second input receiving said new interpolated frame from said frame interpolator means, and said multiplexer receiving at said first output. Outputting one of the accelerated, interlaced video signals and the frame-interpolated frame received at the second input port, and, in the operation, into the existing sequence of frames. 21. The video frame frequency code of claim 20, wherein the newly interpolated frame added increases the frame frequency of the signal. Converta. 22. The first of the standard interlaced video signals.
The frame frequency is about 59.94 fields per second; the higher second frame frequency of the intermediate interlaced video signal is about 60.00 Hz; and the existing A sequence of frames consisting of 1000 video frames, wherein one new interpolated frame is added to each of said 1000 existing frames to increase said first frame frequency. 22. The video frame frequency converter according to claim 21. 23. The buffer memory means stops outputting an existing frame for a period of one frame, and skips one existing frame from the sequence of existing frames, the multiplexer comprising: 23. The video frame frequency converter of claim 22, wherein during the sequence of existing frames, two newly interpolated frames are accepted in place of the one existing frame. 24. The method of claim 1, wherein each of the existing frames comprises two existing fields, and wherein the newly interpolated frame comprises two interpolated video fields. 24. The video frame frequency converter according to claim 23. 25. A pixel calculated by a pixel wherein two interpolated fields are each interpolated using at least two adjacent, interpolated pixels using information constructed among existing pixels. The video frame frequency converter according to claim 24, wherein: 26. A method according to claim 26, further comprising: generating at least one of a set of predefined directions to generate an air-phase direction control signal that is used to perform air-phase complementation in the best direction. Edge direction detector used in a video de-interlacer for detecting the best directions of the past, present, and future received at the input of the direction interpolator Directional interpolator means for performing interpolation on each one of said predefined directions, each one of said predefined directions using a video field signal of Device means for outputting an interpolated signal for each pixel and consisting of signals in said predefined direction; and for interpolation using the interpolated signal from said direction interpolator means. The most Edge direction detecting apparatus characterized by comprising an edge direction selector for selecting the direction of at least one. 27. The past, present and future video field signal is a luminance video signal, and is processed by an edge direction calculator and a filter means for removing noise and enhanced horizontal edges of said interpolated video signal. Wherein the edge direction calculator and the filter means receive an interpolated signal at an input port of the filter means, and output a corrected interpolated signal to the edge direction selector means; Horizontal and vertical detector means for detecting a high frequency region in the resulting signal, said horizontal and vertical high frequency detector receiving said interpolated signal at the input of said device. Outputting, to the edge direction selector means, a high-frequency information signal indicating the high-frequency region in each of the predefined directions, and Regions should be used as possible sources of errors in the edge direction estimation process, edge binary means for filtering and enhancing the output by the edge detector means, and also for interpolation 27. The edge direction detecting device according to claim 26, further comprising direction determining means for selecting one best direction. 28. The pre-defined direction for interpolating the video signal,
Approximately ± 4 degrees, ± 7 degrees, ± 30 degrees, ± 45 degrees and + in a floating reference system having an origin at the pixel to be interpolated and having a first axis oriented horizontally.
27. The edge direction detecting device according to claim 26, wherein an angle of 90 degrees is formed. 29. The directional interpolator means: ## EQU1 ## In the above equation, SST is a first stored luminance signal associated with each of the directions, A ± i is a pixel to be interpolated on a current field line containing the pixel to be interpolated. The values ± i, A ′ ± i of the pixels in the reference system having an origin are the values ± i, B0 of the pixels of the existing field in the system adjacent to the pixel to be interpolated. , The pixel value of the next field to be interpolated correspondingly to the pixel, C0 is the pixel value of the previous field corresponding to the pixel to be interpolated, and D ± i is the pixel value of the next field to be interpolated. In said system having an origin at the pixel to be interpolated on the existing field line containing the pixel, the value of the pixel ± i, D '± i, is the next field line The value of the pixel ± i in the reference system having the origin at the next pixel of the existing line in the next line of the next field, E ± i is the value of the previous field containing the previous field pixel The numerical value of the pixel ± i in the reference system having an origin at the previous field pixel, corresponding to the pixel to be spatially interpolated of the adjacent previously existing previous field, E ′ ± i Is the numerical value of the pixel ± i in the reference system having the origin at the previous field of the existing line of the existing field of the previous field adjacent to the previous field line, F0 is adjacent to pixel A0 at 90 degrees in direction The existing pixel value of the current field, F'0, is the existing pixel value of the current field, adjacent to the pixel corresponding to A'0 in the direction 90 degrees, G Is the existing pixel value of the future field adjacent to D0 in the +90 degree direction, G'0 is the existing pixel value of the future field adjacent to D0 in the -90 degree direction, H0 is , The existing pixel value of the past field adjacent to D0 in the direction of +90 degrees, and H'0 is the existing pixel value of the future field adjacent to E'0 in the direction of 90 degrees. 27. The edge direction detecting device according to claim 26, wherein an interpolated signal defined by the following is output. 30. A quality enhancing converter device for converting a standard interlaced video signal into a progressive HDTV video signal, comprising: converting a standard interlaced video signal having a first frame frequency to a higher second frame frequency. A converter for converting the signal to an intermediate interlaced video signal, the video to frame frequency converter receiving an input of the interlaced standard video signal; and Means for generating an accelerated interlaced video signal, the accelerated video signal having the higher second frame frequency and the interlaced standard video signal. Complement existing existing fields Frame interpolator means for generating at least two new interpolated fields, and inserting said at least two new interpolated fields into said sequence of existing field frames to produce said higher At a second frequency, inserting means for outputting the sequence of existing frames consisting of the interpolated fields as the intermediate interlaced video signal; and analyzing the standard video signal and Frame insertion detector means for detecting an instant and also for adding at least one new frame into said sequence of existing frames of a standard video signal, said frame insertion detector means comprising: When one moment is detected, one insertion frame control signal is generated. And at least one of the best directions from a set of predefined directions to generate the air-phase directional control signal used to perform the air-phase complement in the best direction. Edge direction detection device to be used in a video non-interlacing device receiving said intermediate interlaced video signal at an input to detect the direction of the edge direction detection device. , For each pixel, one of the predefined directions using the past, present and future video field signals consisting of signals interpolated for each of the predefined directions. Each comprising a directional interpolator means for performing interpolation and using said interpolated signal received from said directional interpolator means to select said at least best direction. Quality improvement converter device comprising an edge direction selector means for performing the operation. 31. A past, present and future video field signal,
A luminance video signal, and said edge direction detection and detection device further comprises a direction calculator and filter means for removing noise and enhancing a horizontal edge of said interpolated video signal, said edge direction calculator and a filter. Means for receiving the interpolated signal at the input of the means, and then outputting a corrected interpolated signal to edge direction selector means; and a high frequency in the interpolated signal. Horizontal and vertical high frequency detection device means for detecting an area, wherein said horizontal and vertical high frequency detection device receives said interpolated signal at the garbage of said device, Means for outputting a high-frequency information signal indicating a high-frequency area for each one of the directions defined in advance, wherein the high-frequency area has an edge direction by the operation. Possible errors in the outgoing evaluation process; edge binary selector means for filtering and enhancing the output of the result in said edge direction selector means; and 1 to be used for interpolation. 31. The quality improving converter device according to claim 30, further comprising direction determining means for selecting a best direction of the book. 32. A pre-defined direction for interpolating said video signal,
Approximately ± 4 degrees, ± 7 degrees, ± 30 degrees, ± 45 degrees and + in a floating reference system having an origin at the pixel to be interpolated and having a first axis oriented horizontally.
Forming an angle of 90 degrees, and wherein said direction interpolator means outputs said interpolated signal, said signal being represented by the following equation: Where SST is the first stored luminance signal associated with each of the directions, A ± i is the interpolated value of the current field line containing the pixel to be interpolated. The numerical value ± i, A ′ ± i of the pixel in the reference system having the origin at the pixel is the numerical value ± of the pixel of the existing field in the system adjacent to the pixel to be interpolated. i, B0 are the pixel values of the next field to be interpolated correspondingly to the pixel, C0 is the pixel value of the previous field corresponding to the pixel to be interpolated, D ± i is the interval In the system having an origin at the pixel to be interpolated on an existing field line containing the pixel to be processed, the value of pixel ± i, D ′ ± i, is: The value of the pixel ± i in the reference system having the origin at the next pixel of the existing line in the next field of the next field adjacent to the field line, E ± i is the value of the previous field containing the previous field pixel The value of the pixel ± i in the reference system having the origin at the previous field pixel, corresponding to the pixel to be interpolated in the previous field adjacent to the field to be spatially interpolated, E ' ± i is the numerical value of the pixel ± i in the reference system having the origin at the previous field of the existing line of the previous field adjacent to the previous field line, F0 is the pixel A0 at 90 degrees in direction The existing pixel value of the adjacent current field, F'0, is the existing pixel value of the current field adjacent to the pixel corresponding to A'0 in the direction 90 degrees. G0 is the existing pixel value of the future field adjacent to D0 in the +90 degree direction, and G'0 is the existing pixel of the future field adjacent to D0 in the -90 degree direction. The value H0 is the existing pixel value of the past field adjacent to D0 in the -90 degree direction, and H'0 is the existing pixel value of the future field adjacent to E'0 in the 90 degree direction. 32. The quality improving converter device according to claim 31, wherein an interpolated signal defined as a pixel value is output.