JPH06189190A - Film frame / video field conversion system and its method - Google Patents

Abstract

PURPOSE: To prevent the degradation of picture quality caused by conversion by performing the conversion while applying a proper weight value to an adjacent film frame field, based on the temporal phase relationship of film frame and video field. CONSTITUTION: The digital video signal of film frame at the rate of 24 frames per second is separated into odd-number and even-number frame fields and afterwards, they are converted to digital video field signals at the interval of two lines with the rate of 60 fields per second by a field rate converting device based on the temporal phase relationship of film frame and video field. Namely, as a weight value K, 0.6, 0.4, 0.2 and 0.8 are sequentially multiplied to frames Fx ,o and Fx ,e separated from five film frames F1 -F5 into odd and even numbers, and converted to video fields f1 -f10 through an adder. This is repeated as a unit conversion cycle. Thus, asynchronous or nonuniform conversion is eliminated and in conversion, the degradation of picture quality can be prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a film / video apparatus and method, and more particularly, to video information recorded in a film format (Video).
information is the temporal correlation between the film frame and the video field.
The present invention relates to a system and method for converting video information into a video format (Video format) by using a relation ship.

[0002]

BACKGROUND OF THE INVENTION One of the main sources of video signals is a movie recorded at a rate of 24 frames per second. For example, conversion of such video for TV broadcasting from a film format recorded at a rate of 24 frames per second to a video format of 60 field rates per second typically involves storage tube scanning (st).
various film-scan methods such as line tube scanning using frame tube and digital sequencing. Despite such film-scan method differences, the most commonly used film-video feel ratio conversion techniques are so-called 3: 2.
It is a full-down method. In the 3: 2 full down method, three video fields are scanned from one film frame, and then two video fields are scanned in the film frame. In other words, the two 1 / s
A 24-second film frame will create five 1 / 60-second video fields that will match exactly in time.

FIG. 4 illustrates a prior art conversion technique for converting a film format to a video format in a 3: 2 full down method. Video images are recorded on film at a rate of 24 frames per second. Each video frame is held for 1/30 seconds and each occupies 1/60 seconds in an odd line field.
d) and an even line field.

FIG. 4 shows ten video fields formed from four film frames. Specifically, film frame 1 is scanned with odd film F _1,0 and even field F _2,0 to produce three video fields. That is, the video field 1 is generated from the odd frame-field F _1,0 , the video field 2 is generated from the even frame-field F _{1, e} , and the video field 3 is further generated from the odd field F _1,0.
Is generated from. This is referred to herein as Process 1. The next two video fields consist of the film frame 2 as follows. That is, video field 4 is generated from even frame-field F _{2, e} and video field 5 is generated from odd frame-field F _2,0 . This is process 2. The two fields from film 3 produce the next three video fields. Video field 6 is an even frame-
The field F _{3, e} , the video field 7 is generated from the odd frame-field F _3,0 , and the video field 8 is generated from the even frame-field F _{3, e} . In this specification,
This is called process 3. Finally, the video fields 9 and 10 consist of the film frame 4. Video field 9 is generated from odd frame-field F _4,0 and video field 10 is generated from even frame-field F _{4, e} .

It can be seen that the sequence of processes 1 through 4 described above is repeated to produce a continuous video field. In other words, process 4 is followed by process 1 and the sequence is repeated.

As can be seen from FIG. 4, in the conventional method of converting film frames to video fields using the 3: 2 full-down technique, several film frame-fields, for example video fields with different F _1,0. , For example, repeated for field 1 and field 3. Also, film frames and video fields are desynchronized. That is, each film frame (which is 1/24 seconds) repeatedly produces three or two video fields (each with the same 1/60 second period). Due to such an abnormality, the image quality and definition are poor.

Accordingly, it is an object of the present invention to provide an improved method and system for converting a film frame into a video field using the temporal correlation between the film frame and the video field. .

[0008]

SUMMARY OF THE INVENTION In accordance with the present invention, a video signal consisting of a film frame having any of the frames is converted to a video signal of an interlaced video field having a higher field rate. , A new method is provided, wherein the method stores (or steps) the film frames, generating one odd frame field and one even frame field from each of the film frames, and the odd and even frame fields. Based on the temporal correlation between a film frame or the like and a video field or the like.

The present invention also provides a new system for converting a video signal composed of film frames having frames into a video signal of a remote video field having a field rate higher than said frame rate, said system comprising: Means for storing the film frames, means for generating one odd frame field and one even frame field from each of the film frames and between the odd and even frame fields such as a film frame and a video field Means for converting to the alternate video field based on the temporal correlation of the.

According to a first feature of the invention, the video fields are generated from film in such a way that the same video field does not repeat.

According to another feature of the invention, the film frame / video film conversion is based on a temporal correlation between a film frame or the like and a video field or the like, with an appropriate weighting value for adjacent film frame fields. Is given.

[0012]

1 is a schematic block diagram of a conversion system for converting a 24 film frame rate per second to a 60 video field rate per second according to a preferred embodiment of the present invention.

The digital video signal in the film format is stored in a semiconductor memory device or a general frame memory (memory device) 100 formed similarly thereto. Preferably, a frame memory (storage device) 100 having a storage capacity capable of storing two frames is
Store film frames at a frame rate of 24 frames per second. The recording of the frame signal in the frame memory 100 is controlled and updated so that the stored frame signal can hold the necessary input signal to the frame / field conversion block 110. The digital signal stored in the frame memory 100 is input to a frame / field conversion block 110, and the frame / field conversion block 110 selectively reads a video signal having a frame format and separates odd and even frames / fields from the read video signal. And generate it. A field rate conversion block 120 converts the generated 24 frame per second odd and even frames / fields to a 60 field per second alternate digital video field signal, which conversion is between the film frame and the video field. It is performed based on temporal correlation.

The conversion of film frames into video fields in the field rate conversion block 120 is illustrated in FIG.
And described in detail with reference to FIG.

FIG. 2 shows the temporal correlation between a film frame at a rate of 24 frames per second and an alternating video field at a rate of 60 field per second. F _x is x
Represents the th film frame, where x is a positive integer. F _{x, 0} is an odd frame / field which is formed by the odd lines of F _x, F _{x, e} is an even frame / field which is formed by the even lines of F _x. Note that the odd and even frame fields F _{x, 0} and F _{x, e} are filmed in frame F _x . In other words, it should be noted that they occur at the same time with a cycle of 1/24 seconds. f _y represents an alternating odd or even video field, where y is a positive integer. 5/60 seconds and 2/24
Since the seconds represent the same time, it can be seen that there are five video fields for every two films.

In the preferred embodiment of the invention, the video field f _y (eg, f ₆ or f ₁₁ ) is every 5th.
Every / 60 seconds, when f _y matches one even or odd frame field (eg, F _{3, e} or F _{5, e} ), it is generated directly from each of them. That is, the odd video field f ₁ consists of the odd frames / fields F _{1,0 of} the first film frame F ₁ and f ₆ is the third
Even frame / field F of film frame F _3
It is performed by the formula consisting of _{3, e} .

[0017] f _2, f 3 in FIG. _2, f _4, as in _{f 5, f 7, f 8} , f 9 and f _10, such as a film frame and the asynchronous video field parity, such as those fields 2 adjacent film frame-fields with .gamma. For example, the even video field f ₂ has two even frame fields F 2.
Generated from _{1, e} and F _{2, e} , the odd video field f ₃ has two odd frame fields F _1,0 and F
Generated from _2,0 .

In accordance with a preferred embodiment of the present invention, the weighting value is determined to be inversely proportional to the ratio of the time intervals between two frame fields, etc. and the video field formed therefrom. For example, the time interval between f ₂ and _{F 1, e T (f 2} , F 1, e) and f ₂ and F _2, the time interval between the _e T of (f _2, F _{2, e)} Since the ratio is 4: 6, f ₂ is F
_It can be obtained as the sum of the value obtained by multiplying _{1, e} by 0.6 and the value obtained by multiplying F _{2, e} by the weighted value 0.4. As can be seen from the above, larger weights are applied to closer frame fields according to the preferred embodiment. In other words, the preferred embodiment gives greater weight to closer frame fields. Similarly, f ₅ is 0.4
_Obtained by adding x F _2,0 and 0.6 x F _3,0 .

Referring to FIG. 3, there is shown a procedure for converting film frames at a rate of 24 frames per second into alternate video signals at a rate of 60 fields per second. The digitized film frame signal is stored in frame memory (storage device) 100 shown in FIG. 1 in frame units. Since each video field consists of one or two closest film frames, the frame memory (storage device) 100 of FIG. 1 preferably has a storage capacity of two frames. Each frame F _x stored in the frame memory 100 is converted into one odd field F _{x, 0} by the frame / field conversion block 110 of FIG.
And an even field F _{x, e} . These film frame fields and the like are converted to video fields by the field rate conversion block 120 according to the preferred embodiment detailed with reference to FIG.

FIG. 3 shows one of five film frames.
The conversion to zero alternate video fields is shown.
As described below, this is a repeating unit conversion cycle that is repeated in order to completely convert an entire film frame at a rate of 24 frames per second into an alternating video field at a rate of 60 field per second.

In the present invention, the field rate conversion is performed as follows through the standard multiplier and adder shown in FIG.

F ₁ = F _1,0 f ₂ = k ₁ F _{1, e} + k ₂ F _{2, e} f ₃ = k ₃ F _1,0 + k ₄ F _2,0 f ₄ = k ₄ F _{2, e} + k ₃ F _{3, e} f ₅ = k ₂ F _2,0 + k ₁ F _3,0 f ₆ = F _{3, e} f ₇ = k ₁ F _3,0 + k ₂ F _4,0 f ₈ = k ₃ F _{3, e} + k ₄ F _{4, e} f ₉ = k ₄ F _4,0 + k ₃ F _5,0 f ₁₀ = k ₂ F _{4, e} + k ₁ F _{5, e} where k ₁ , k ₂ , k ₃ and k ₄ is a weighted value, k ₁ = 0.6, k ₂ = 0.4, k ₃ = 0.2, k ₄
= 0.8.

These k-values and the like are determined by the time interval T (F _i , F
_{i +1)} and _{T (f i, f i +1} ) and the like (also,
The time interval between each even frame that is 1/120 second and its nearest video field, eg, T
It is determined directly (from f ₃ , F ₂ ) and T (F ₂ , f ₄ ). Here, i is a constant, and these T values are 1/24 and 1/60, respectively, as shown in FIG.
Seconds.

Referring to FIG. 4, it can be seen that the sequence of processes is repeated for the generation of consecutive odd and even video fields, as in the conventional 3: 2 full-down method described above. The time correlation between the film frame field and the video field is not every 2 film frames (or 2/24 seconds) or 5 video fields (or 5/60 seconds) but every time by the special field rate conversion method of the present invention. 4 film frames, or 10
Care must be taken to repeat every video field.

In the preferred embodiment of the present invention, a weighting value that is inversely proportional to the time interval between any video field and its two adjacent film frames is obtained by other methods. For example, if the weights obtained by other methods are efficient for playing better video footage, then those weights will have any identical value, etc., regardless of time intervals. It can also be a value that is inversely proportional to the square of the ratio of the time intervals.

Further, the temporal correlation between the film frame and the video field can be calculated as the first video field f.
They can also be coordinated with each other by placing ₁ in other temporal positions. For example, f ₁ could be located between film frames F ₁ and F ₂ and the weight value determined thereby.

The present invention is also applicable to converting video information from any media to other media having a variety of other criteria and characteristics.

[0028]

The new conversion method according to the present invention is 2 seconds per second.
Converting four film frames into 60 video fields per second, where conversion is done by giving appropriate weights to adjacent film frame fields based on their temporal correlation with film frames. , Flatten the 6 Hz difference between 24 film frames and 30 video frames (or 60 video fields) to solve the problem of image degradation.

In other words, unlike conventional systems which use the 3: 2 full down conversion method, according to the present invention, no two video fields have the same data from one frame field, and There is no asynchronous or non-uniform conversion between the film frame and the video field.

[Brief description of drawings]

1 is a block of the system according to the invention for converting a film frame into an alternating field.

FIG. 2 is a diagram showing a temporal correlation between a film frame having a rate of 24 frames per second and a video field having a rate of 60 field per second.

FIG. 3 is an illustration of a procedure in accordance with the present invention for converting a film frame at a rate of 24 frames per second into an alternating video field at a rate of 60 field per second.

FIG. 4 is an illustration of a conventional 3: 2 full-down method for converting film frames at a rate of 24 frames per second to alternating video fields at a rate of 60 fields per second.

[Explanation of symbols]

F _i film frame F _{i, 0} , F _{i, e} film frame field 100 frame memory 110 frame / field conversion block 120 field rate conversion block f _y video field

Claims (12)

[Claims] 1. A method for converting a video signal provided in a film frame at any frame rate into a video signal in a remote video field having a field rate greater than the frame rate, the method comprising: the film frame. In the storage means, generating one odd frame field and one even frame field from each film frame retrieved from the storage means, and the video field. Converting the odd and even frame fields into the alternating video fields based on the temporal correlation between the film frames and the video fields so as not to repeat the same video signal generated from the odd and even frame fields. A method that includes the steps to . 2. The method of claim 1, wherein each said alternate video field comprises a frame field having the same parity as one or two temporally closest video fields therefrom. 3. When each video field consists of the two frame fields of the same parity,
3. A weight value greater than the weight value assigned to another frame field is assigned to a frame field that is temporally closer to each said video field.
The method described. 4. The method of claim 3, wherein each video field comprises a frame field having the same parity of the frame when synchronized with any frame. 5. The method of claim 2, wherein the frame rate is 24 frames per second and the field rate is 60 fields per second. 6. The method of claim 5, wherein the video field is generated by repeating the following unit order. f ₁ = F _1,0 f ₂ = k ₁ F _{1, e} + k ₂ F _{2, e} f ₃ = k ₃ F _1,0 + k ₄ F _2,0 f ₄ = k ₄ F _{2, e} + k ₃ F _{3 , e} f ₅ = k ₂ F _2,0 + k ₁ F _3,0 f ₆ = F _{3, e} f ₇ = k ₁ F _3,0 + k ₂ F _4,0 f ₈ = k ₃ F _{3, e} + k ₄ F _{4, e} f ₉ = k ₄ F _4,0 + k ₃ F _5,0 f ₁₀ = k ₂ F _{4, e} + k ₁ F _{5, e} where f ₁ , f ₂ ... f ₁₀ are weighted values , K ₁ =
Using 0.6, k ₂ = 0.4, k ₃ = 0.2, k ₄ = 0.8, odd frame field F _{i, 0} and even frame field F _{i, e} (where i is 1 Is a constant video field of 5 to 5). 7. A system for converting a video signal provided to a film frame at any frame rate to a video signal in a remote video field having a field rate greater than said frame rate, said system comprising: said film frame. Means for storing one odd frame field and one even frame field from each of the film frames retrieved from the storage means, and any one of the video fields. In order not to repeat the same video signal generated from the odd and even frame fields, the odd and even frame fields are divided into the alternate video fields etc. based on the temporal correlation between the film frame and the video field. To convert And a system including. 8. The system of claim 7, wherein each said alternate video field comprises a frame field having the same parity as one or two temporally closest ones therefrom. 9. When each of the video fields is composed of the two frame fields of the same parity, etc., a weight value greater than a weight value assigned to another frame field is temporally greater in each of the video fields. The system of claim 8, wherein the system is assigned to near frame fields. 10. The system of claim 8, wherein each video field comprises a frame field having the same parity of the frame when synchronized with any frame. 11. The system of claim 8, wherein the frame rate is 24 frames per second and the field rate is 60 fields per second. 12. The system of claim 11, wherein the video field is generated by repeating the following unit order. f ₁ = F _1,0 f ₂ = k ₁ F _{1, e} + k ₂ F _{2, e} f ₃ = k ₃ F _1,0 + k ₄ F _2,0 f ₄ = k ₄ F _{2, e} + k ₃ F _{3 , e} f ₅ = k ₂ F _2,0 + k ₁ F _3,0 f ₆ = F _{3, e} f ₇ = k ₁ F _3,0 + k ₂ F _4,0 f ₈ = k ₃ F _{3, e} + k ₄ F _{4, e} f ₉ = k ₄ F _4,0 + k ₃ F _5,0 f ₁₀ = k ₂ F _{4, e} + k ₁ F _{5, e} where f ₁ , f ₂ ... f ₁₀ are weighted values , K ₁ =
Using 0.6, k ₂ = 0.4, k ₃ = 0.2, k ₄ = 0.8, odd frame field F _{i, 0} and even frame field F _{i, e} (where i is 1 Is a constant video field of 5 to 5).