JP4835234B2 - Moving picture conversion apparatus, program thereof, and moving picture conversion method - Google Patents

Description

The present invention relates to a moving image conversion apparatus, a program thereof, and a moving image conversion method, and more particularly to a moving image conversion apparatus that performs frame interpolation in order to convert a moving image to a high frame rate, a program thereof, and a moving image conversion method.

In the conventional method of converting a low frame rate moving image to a high frame rate, the same frame image is displayed a plurality of times, and frame interpolation is performed by filling in missing frames, thereby converting to a high frame rate. Here, this interpolation method is called interpolation by copying.

For example, in the case of conversion from 30 fps (frame per second) to 60 fps, (
As shown in a), an image of the original frame is copied and displayed again to form an interpolation frame.

In addition, when the motion of the entire image from the previous image to the current image is calculated and converted to a frame rate of n times, an image obtained by shifting the previous image by 1 / n of the motion of the entire image is displayed. In some cases, frame interpolation is performed to convert the frame rate to a high frame rate (for example, Patent Document 1).
, See Patent Document 2). Here, this interpolation method is called interpolation by parallel movement.

For example, in the case of conversion from 30 fps to 60 fps, as shown in (b) of FIG. 5, Val, which is the movement of the entire image between the original frames, is calculated and converted to a double frame rate. The interpolated frame is an image obtained by shifting the above image by Val / 2. Thereby, a smooth movement is displayed as compared with FIG.
JP 60-28392 A JP 2000-165821 A

However, in the interpolation method based on parallel movement shown in Patent Document 1, if there is a part that moves differently from the whole in the image, there is a problem that the image of the part deteriorates. For example,
As shown in FIG. 6B, when the large portion P1 in the image moves to the right while the small portion P2 is stationary, the movement of the entire image is close to the movement of the portion P1. Therefore, the motion of the part P1 becomes smooth by interpolating the frame. However, the originally stationary part P2 moves to the right from the original position in the interpolated frame and moves to the left in the original frame to return to the original position, resulting in a jagged movement and deterioration of the image. In this case, in the interpolation method by copying, as shown in FIG. 6A, the portion P2 remains stationary, so that the image does not deteriorate, but the portion P1 remains stationary between the original frame and the interpolation frame. There is a problem that the movement is not smooth because it moves only from the interpolated frame to the next original frame.

The present invention has been made in view of such circumstances, and an object of the present invention is to provide a moving image conversion apparatus capable of obtaining a smooth moving image even after frame rate conversion while suppressing image deterioration, a program thereof, and a moving image It is to provide a conversion method.

The present invention has been made to solve the above-described problems, and the moving image conversion apparatus according to the present invention is configured such that the frame rate of the input image is Ri and the frame rate of the output image is Ro, and the frame interpolation enlargement ratio. In the moving image conversion apparatus that performs frame interpolation of Ex = Ro / Ri,
An overall calculation means for calculating an overall motion vector Vall, which is a motion vector from the entire image one frame before to the entire current image, for the input image;
Area calculation means for calculating a motion vector Vmn for each area (n represents an nth area), which is a motion vector from the previous image to the current image, for each area obtained by dividing the image into a plurality of areas; ,
As a result of comparing the overall motion vector Vall calculated by the overall calculation unit and the value Vall (Ex-1) / Ex calculated from the frame interpolation magnification rate Ex with the motion vector Vmn for each region calculated by the region calculation unit. Accordingly, as a frame interpolation method, an interpolation method selection means for selecting from a method of interpolating the image of the previous frame by an image shifted by Val / Ex and a method of interpolating by the image of the previous frame. Prepared ,
The interpolation method selection means sets the horizontal and vertical components of the entire motion vector Vall as Vall_x and Vall_y, respectively, and sets the horizontal and vertical components of the region-specific motion vector Vmn as Vmn_x and Vmn_y, respectively, in a predetermined number α of regions.
(Vmn_x ≧ Vall_x (Ex−1) / Ex ≧ 0, or Vmn_x ≦ Vall_x (Ex−1) / Ex <0) and
(Vmn_y ≧ Vall_y (Ex−1) / Ex ≧ 0 or Vmn_y ≦ Vall_y (Ex−1) / Ex <0)
Is time holds, the one frame preceding image select and interpolated by the image shifted by Vall / Ex, at other times, characterized by you to select the interpolated by the preceding frame image.

The program of the present invention is a computer,
A moving image conversion means for performing frame interpolation with a frame interpolation magnification rate Ex = Ro / Ri, where Ri is the frame rate of the input image and Ro is the frame rate of the output image,
An overall calculation means for calculating an overall motion vector Vall, which is a motion vector from the entire image one frame before to the entire current image, for the input image;
Area for calculating a motion vector Vmn (n represents the nth area) for each area, which is a motion vector from the previous frame image to the current image, for each area obtained by dividing the current image into a plurality of areas. A calculation means;
As a result of comparing the overall motion vector Vall calculated by the overall calculation means and the value Vall (Ex-1) / Ex calculated from the frame interpolation magnification rate Ex with the motion vector Vmn for each area calculated by the area calculation means. Accordingly, as a frame interpolation method, an interpolation method selecting means for selecting from a method of interpolating the image of the previous frame by an image shifted by Val / Ex and a method of interpolating by the image of the previous frame. Prepared ,
The interpolation method selection means sets the horizontal and vertical components of the entire motion vector Vall as Vall_x and Vall_y, respectively, and sets the horizontal and vertical components of the region-specific motion vector Vmn as Vmn_x and Vmn_y, respectively, in a predetermined number α of regions.
(Vmn_x ≧ Vall_x (Ex−1) / Ex ≧ 0, or Vmn_x ≦ Vall_x (Ex−1) / Ex <0) and
(Vmn_y ≧ Vall_y (Ex−1) / Ex ≧ 0 or Vmn_y ≦ Vall_y (Ex−1) / Ex <0)
Is time holds, the one frame preceding image select and interpolated by the image shifted by Vall / Ex, at other times, characterized by you to select the interpolated by the preceding frame image.

Also, the moving image conversion method of the present invention is a moving image conversion apparatus that performs frame interpolation with a frame interpolation enlargement ratio Ex = Ro / Ri, where Ri is the frame rate of the input image and Ro is the frame rate of the output image. In the moving image conversion method in
A first process of calculating an overall motion vector Vall, which is a motion vector from the entire image one frame before to the entire current image, for the input image;
For each region obtained by dividing the current image into a plurality of regions, a motion vector Vmn (n represents the nth region) for each region, which is a motion vector from the previous image to the current image, is calculated. 2 process,
A value Vall (Ex-1) / Ex calculated from the overall motion vector Vall calculated in the first process and the frame interpolation magnification rate Ex, and an area-specific motion vector Vmn calculated in the second process. According to the comparison result, a frame interpolation method is selected from among a method of interpolating the image of the previous frame by an image shifted by Val / Ex and a method of interpolating by the image of the previous frame. And the process
Equipped with a,
In the third step, the horizontal and vertical components of the entire motion vector Vall are set as Vall_x and Vall_y, respectively, and the horizontal and vertical components of the motion vector Vmn for each region are set as Vmn_x and Vmn_y, respectively, in a predetermined number α of regions,
(Vmn_x ≧ Vall_x (Ex−1) / Ex ≧ 0, or Vmn_x ≦ Vall_x (Ex−1) / Ex <0) and
(Vmn_y ≧ Vall_y (Ex−1) / Ex ≧ 0 or Vmn_y ≦ Vall_y (Ex−1) / Ex <0)
Is time holds, the one frame preceding image select and interpolated by the image shifted by Vall / Ex, at other times, characterized by you to select the interpolated by the preceding frame image.

As a result, the moving image conversion apparatus (or program or moving image conversion method) of the present invention can determine the frame interpolation method by determining whether the motion for each region is different from the motion of the entire image. An image with smooth motion can be obtained even after frame rate conversion.

Further, the moving image conversion apparatus of the present invention is the above-described moving image conversion apparatus, wherein the interpolation method selection means sets the horizontal component and the vertical component of the overall motion vector Vall as Val_x,
Let Vall_y be the horizontal component and vertical component of the motion vector Vmn for each region, Vm
As n_x and Vmn_y, in a predetermined number α of regions,
(Vmn_x ≧ Vall_x (Ex−1) / Ex ≧ 0 or Vmn_x ≦ Vall
_X (Ex−1) / Ex <0) and (Vmn_y ≧ Vall_y (Ex−1) / Ex ≧ 0 or Vmn_y ≦ Vall
_Y (Ex-1) / Ex <0)
When the above holds, it is selected that the image one frame before is interpolated by an image shifted by Val / Ex, and otherwise, it is selected that the image is interpolated by the image one frame before.

As a result, the moving image conversion apparatus of the present invention can detect a region where the image deteriorates when the image of the previous frame is interpolated by an image shifted by Val / Ex, so that the predetermined number α is set to an appropriate value. As a result, it is possible to obtain a smooth motion image after frame rate conversion while suppressing image degradation.

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1A is a schematic block diagram showing a configuration of a moving image conversion apparatus 100 according to an embodiment of the present invention. In the present embodiment, the moving image conversion apparatus 100 converts a 30 fps moving image to 60 fps and outputs the converted image. Reference numeral 101 denotes an input I / F (Interface) that receives an input of a moving image. 102 is
A RAM (Random Access Memory) used as a work area. Reference numeral 103 denotes a CPU (Central Processing Unit) that controls the operation of the moving image conversion apparatus 100. As shown in FIG. 1B, the CPU 103 includes an overall calculation processing unit (overall calculation unit) 110, an area calculation processing unit (region calculation unit) 111, an interpolation method determination processing unit (interpolation method selection unit) 112,
A translation interpolation processing unit 113 and a copy interpolation processing unit 114 are provided. Details of the CPU 103 will be described in the description of the operation. A ROM (Read Only Memory) 104 stores a program and data for the CPU 103 to operate. 105 is an input I / F1
01 is a hard disk that temporarily stores a moving image received and a moving image obtained by converting the frame rate. Reference numeral 106 denotes an output I / F that outputs the converted moving image. Reference numeral 107 denotes an input I / F 101, RAM 102, CPU 103, ROM 104, hard disk 105,
This is a bus for connecting the output I / F 107.

A video camera 200 captures a subject and outputs a 30 fps moving image. 21
Reference numeral 0 denotes a moving image reproduction device that reads a moving image recorded on a video tape, a DVD (Digital Versatile Disk), or the like and outputs a 30 fps moving image. Reference numeral 300 denotes a projector that receives and projects a moving image of 60 fps. A display 310 receives and displays a moving image of 60 fps.

Next, the operation of the moving image conversion apparatus 100 will be described. When the video camera 200 or the video playback device 210 outputs a moving image of 30 fps, the input I / F 101 receives this,
Records in the hard disk 105. The CPU 103 reads out the frames of the moving image recorded on the hard disk 105 in time order by the program recorded in the ROM 104, operates as shown in the flowchart of FIG. Perform conversion.

First, the entire calculation processing unit 110 of the CPU 103 calculates the entire motion vector Vall of the entire image between two temporally continuous frames (frame A and frame B in time order) read from the hard disk 105 (S1). ). The method for calculating the motion vector will be described in detail later. Next, the calculation processing unit 111 for each region of the CPU 103 divides the image into regions divided as shown in FIG. 3, and for each region, a motion vector Vmn for each region between the frame A and the frame B (n is (representing the nth region) is calculated (S2). For example, when the area is divided into 7 × 7 = 49 areas as shown in FIG. 3, the area-specific motion vectors Vm1 to Vm49 are calculated for the 1st to 49th areas. Next, the interpolation method determination processing unit 112 of the CPU 103 determines the horizontal component and the vertical component of the entire motion vector Val as Val_x, Val.
l_y, the horizontal component and vertical component of the motion vector Vmn for each region are represented by Vmn_x, Vmn_y
As above, the number N of regions satisfying the following conditions is calculated.

When Vall_x ≧ 0 Vmn_x ≧ Vall_x (EX-1) / EX
When Val_x <0 Vmn_x ≦ Vall_x (EX-1) / EX
And when Vall_y ≧ 0 Vmn_y ≧ Vall_y (EX-1) / EX
When Val_y <0 Vmn_y ≦ Vall_y (EX-1) / EX
Here, EX is a frame interpolation magnification rate.

In this embodiment, the frame interpolation enlargement ratio = output frame rate Ro / input frame rate Ri = 60 fps / 30 fps = 2.
The interpolation method determination processing unit 112 calculates the number N of interpolations between the frame A and the frame B.
Is greater than or equal to a predetermined threshold value α, it is selected to use the translation interpolation processing unit 113 that performs interpolation by translation. If the calculated number N is less than the predetermined threshold value α, a copy interpolation processing unit that performs interpolation by copying If 114 is used, it selects (S3). Next, according to the selection of the interpolation method selection processing unit 112 in step S3, the translation interpolation processing unit 113 or the copy interpolation processing unit 114 generates an interpolation frame, and inserts this interpolation frame between the frame A and the frame B. Then, it is converted into a moving image of a target frame rate (here, 60 fps) (S4). For example, if the translation interpolation processing unit 113 is selected in step S3, the translation interpolation processing 113 is EX = 2 in the present embodiment, and therefore, the image 1 obtained by shifting the frame A by Val / 2.
A frame is an interpolation frame. When converting from 30 fps to 90 fps, EX = 3
Therefore, two images, that is, an image obtained by shifting Frame A by Val / 3 and an image obtained by shifting 2 × Val / 3 are used as interpolation frames. If the copy interpolation processing unit 114 is selected in step S3, the copy interpolation processing unit 114 sets one image obtained by copying the frame A as an interpolation frame.

When the CPU 103 finishes the frame interpolation between the frame A and the frame B, the CPU 103 reads out the frame C whose time sequence in the input moving image is next to the frame B from the hard disk 105, and performs the above-described step S1 from the same as the above. By repeating the above, frame interpolation between the frame B and the frame C is performed. By repeating such an operation, the input moving image is converted into a moving image having a target frame rate (here, 60 fps).

The CPU 103 outputs the moving image converted to the target frame rate (60 fps).
Output via / F106. When the projector 300, the display 310, and the like connected to the output I / F 106 receive the converted moving image, they are displayed.

Note that the output of the converted moving image via the output I / F 106 may be performed while the conversion is being performed successively, or may be performed after the converted moving image is once stored in the hard disk 105.

Next, a conditional expression for selecting a frame interpolation method in the interpolation method determination processing unit 112 will be described. Here, in order to simplify the explanation, it is assumed that the entire motion vector Vall and the region-specific motion vector Vmn are one-dimensional values. Assume that the relationship between the overall motion vector Vall from the frame A to the frame B and the motion vectors Vm1 to Vm4 for each region is as shown in FIG. At this time, if interpolation is performed by translation, frame A is Val.
The frame moved by l / 2 is inserted between the frame A and the frame B as an interpolation frame. Here, paying attention to the n-th region, the subject that was at position 0 in frame A is displayed at the position of Val / 2 in the interpolated frame, and is displayed at the position of Vmn in frame B.

Returning to FIG. 4A, this is applied to each of the first to fourth areas. In the first region, since 0 ≦ Val / 2 <Vm1, the frame A, the interpolation frame, and the frame B always move in the right direction. Even in the second region, 0 ≦ Val / 2 <Vm2, and therefore always moves in the right direction. However, in the third region, 0 ≦ Val / 2> Vm3
Therefore, the subject that once moved to the right in the interpolation frame moves to the left in the frame B, so that the image is deteriorated as a jagged movement. Even in the fourth region, 0 ≦ Vall / 2> V
Since it is m4, it is the same as the third region. That is, when Vall ≧ 0, the condition for preventing the subject in the nth area from moving in a jagged manner is expressed by the following equation (1).

Vall / 2 ≦ Vmn (1)
Here, since the frame interpolation enlargement ratio EX = 2, there is one interpolation frame, Va
The frame A is moved by ll / 2, but when this is generalized and the frame interpolation enlargement ratio EX, the number of interpolation frames is EX-1, and each interpolation frame is Val.
The frame A is moved by l / EX, 2Vall / EX, 3Vall / EX ... Vall (EX-1) / EX. From this, when the formula (1) is generalized, the formula (2) is obtained.

Vall (EX-1) / EX ≦ Vmn (2)
This is equivalent to comparing the motion vector of the last (EX-1) interpolated frame from frame A with the motion vector of frame A to frame B.

When the left side and the right side of equation (2) are interchanged, equation (3) is obtained, which is the interpolation method determination processing unit 11.
This corresponds to the first and third expressions of the conditional expression 2 (step S3).

Vmn ≧ Vall (EX-1) / EX (3)
If Vall <0, as shown in the example of FIG. 4B, the same as described above (4)
This is an expression, which corresponds to the second and fourth expressions of the conditional expression of the interpolation method determination processing unit 112 (step S3).

Vmn ≦ Vall (EX-1) / EX (4)
That is, the interpolation method determination processing unit 112 determines the interpolation method based on whether or not the number of regions where deterioration does not occur when interpolation by parallel movement is performed is equal to or greater than the threshold value α.

Note that the threshold value α may be set to the number of regions (49 in this embodiment) so that deterioration does not occur in all regions.
[Calculation of motion vector]
Whole calculation processing unit 110 (step S1) and area calculation processing unit 111 (step S2)
The method of calculating the motion vector that is to be calculated in () will be described. The calculation method is roughly divided into a block matching method and a gradient method. Here, the block matching method will be described. The block matching method is a method in which a block most similar to a target block is searched from a reference image, and a vector from the target block to the searched block is used as a motion vector. For the evaluation of similarity, the sum of absolute values of differences between corresponding pixels in the block is used.

In this embodiment, for example, when calculating the motion vector of the nth region from frame A to frame B, an arbitrary point within the fixed range from the center of the nth region is the same size as the nth region. A comparison between the block R of the frame B centered at the center and the nth region of the frame A is performed for all within the corresponding range while shifting the position of the block R by one pixel, and the most similar block is detected. .

In the present embodiment, the overall calculation processing unit 110, the region calculation processing unit 111, the interpolation method determination processing unit 112, the parallel movement interpolation processing unit 113, and the copy interpolation processing unit 114 are realized by dedicated hardware. In addition, the memory (RAM102, ROM1
04, the hard disk 105) and the CPU 103.

2 includes a whole calculation processing unit 110, a region-by-region calculation processing unit 111, an interpolation method determination processing unit 112, a parallel movement interpolation processing unit 113, and a copy interpolation processing unit 114, and functions as a moving image conversion unit that performs frame interpolation. Frame interpolation may be performed by recording a program for realizing on a computer-readable recording medium, causing the computer system to read and execute the program recorded on the recording medium. Here, the “computer system” includes an OS and hardware such as peripheral devices.

The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes design and the like within a scope not departing from the gist of the present invention.

It is a block diagram which shows the structure of the moving image converter by one Embodiment of this invention. It is a flowchart explaining operation | movement of CPU103 in the embodiment. It is a figure explaining the whole motion vector Vall and the motion vector Vmn for every area in the embodiment. It is a figure explaining the criterion in the interpolation method determination process 112 in the same embodiment. It is a figure explaining the interpolation by the copy which is the conventional frame interpolation method, and the interpolation by parallel movement. It is a figure explaining the problem of the interpolation by the copy which is the conventional frame interpolation method, and the interpolation by parallel movement.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Moving image converter 101 ... Input I / F
102 ... RAM
103 ... CPU
104 ... ROM
105 ... Hard disk 106 ... Output I / F
DESCRIPTION OF SYMBOLS 107 ... Bus 110 ... Whole calculation process part 111 ... Area calculation process part 112 ... Interpolation method determination process part 113 ... Parallel displacement interpolation process part 114 ... Copy interpolation process part 200 ... Video camera 210 ... Movie reproduction apparatus 300 ... Projector 310 ... display

Claims (3)

In a moving image conversion apparatus that performs frame interpolation with a frame interpolation magnification rate Ex = Ro / Ri, where Ri is the frame rate of the input image and Ro is the frame rate of the output image,
An overall calculation means for calculating an overall motion vector Vall, which is a motion vector from the entire image one frame before to the entire current image, for the input image;
Area calculation means for calculating a motion vector Vmn for each area (n represents an nth area), which is a motion vector from the previous image to the current image, for each area obtained by dividing the image into a plurality of areas; ,
As a result of comparing the overall motion vector Vall calculated by the overall calculation unit and the value Vall (Ex-1) / Ex calculated from the frame interpolation magnification rate Ex with the motion vector Vmn for each region calculated by the region calculation unit. Accordingly, as a frame interpolation method, an interpolation method selection means for selecting from a method of interpolating the image of the previous frame by an image shifted by Val / Ex and a method of interpolating by the image of the previous frame. Prepared ,
The interpolation method selection means sets the horizontal and vertical components of the entire motion vector Vall as Vall_x and Vall_y, respectively, and sets the horizontal and vertical components of the region-specific motion vector Vmn as Vmn_x and Vmn_y, respectively, in a predetermined number α of regions.
(Vmn_x ≧ Vall_x (Ex−1) / Ex ≧ 0, or Vmn_x ≦ Vall_x (Ex−1) / Ex <0) and
(Vmn_y ≧ Vall_y (Ex−1) / Ex ≧ 0 or Vmn_y ≦ Vall_y (Ex−1) / Ex <0)
When the holds, the 1 choose interpolated by the frame image before the image shifted by Vall / Ex, moving images at other times, characterized by you to select the interpolated by the previous frame image Conversion device. Computer
A moving image conversion means for performing frame interpolation with a frame interpolation magnification rate Ex = Ro / Ri, where Ri is the frame rate of the input image and Ro is the frame rate of the output image,
An overall calculation means for calculating an overall motion vector Vall, which is a motion vector from the entire image one frame before to the entire current image, for the input image;
Area for calculating a motion vector Vmn (n represents the nth area) for each area, which is a motion vector from the previous frame image to the current image, for each area obtained by dividing the current image into a plurality of areas. A calculation means;
As a result of comparing the overall motion vector Vall calculated by the overall calculation means and the value Vall (Ex-1) / Ex calculated from the frame interpolation magnification rate Ex with the motion vector Vmn for each area calculated by the area calculation means. Accordingly, as a frame interpolation method, an interpolation method selecting means for selecting from a method of interpolating the image of the previous frame by an image shifted by Val / Ex and a method of interpolating by the image of the previous frame. Prepared ,
The interpolation method selection means sets the horizontal and vertical components of the entire motion vector Vall as Vall_x and Vall_y, respectively, and sets the horizontal and vertical components of the region-specific motion vector Vmn as Vmn_x and Vmn_y, respectively, in a predetermined number α of regions.
(Vmn_x ≧ Vall_x (Ex−1) / Ex ≧ 0, or Vmn_x ≦ Vall_x (Ex−1) / Ex <0) and
(Vmn_y ≧ Vall_y (Ex−1) / Ex ≧ 0 or Vmn_y ≦ Vall_y (Ex−1) / Ex <0)
When the holds, the 1 choose interpolated by the frame image before the image shifted by Vall / Ex, moving images at other times, characterized by you to select the interpolated by the previous frame image A program for functioning as a conversion means. In the moving image conversion method in the moving image conversion apparatus that performs frame interpolation with the frame interpolation magnification rate Ex = Ro / Ri, where Ri is the frame rate of the input image and Ro is the frame rate of the output image,
A first process of calculating an overall motion vector Vall, which is a motion vector from the entire image one frame before to the entire current image, for the input image;
For each region obtained by dividing the current image into a plurality of regions, a motion vector Vmn (n represents the nth region) for each region, which is a motion vector from the previous image to the current image, is calculated. 2 process,
A value Vall (Ex-1) / Ex calculated from the overall motion vector Vall calculated in the first process and the frame interpolation magnification rate Ex, and an area-specific motion vector Vmn calculated in the second process. According to the comparison result, a frame interpolation method is selected from among a method of interpolating the image of the previous frame by an image shifted by Val / Ex and a method of interpolating by the image of the previous frame. And the process
Equipped with a,
In the third step, the horizontal and vertical components of the entire motion vector Vall are set as Vall_x and Vall_y, respectively, and the horizontal and vertical components of the motion vector Vmn for each region are set as Vmn_x and Vmn_y, respectively, in a predetermined number α of regions,
(Vmn_x ≧ Vall_x (Ex−1) / Ex ≧ 0, or Vmn_x ≦ Vall_x (Ex−1) / Ex <0) and
(Vmn_y ≧ Vall_y (Ex−1) / Ex ≧ 0 or Vmn_y ≦ Vall_y (Ex−1) / Ex <0)
When the holds, the 1 choose interpolated by the frame image before the image shifted by Vall / Ex, moving images at other times, characterized by you to select the interpolated by the previous frame image Conversion method.

Images