JPH1198534A - Method for generating three-dimensional image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an impressive three-dimensional image with an excellent stereoscopic modification by detecting the state of a two-dimensional image, based on the motion vector of the two-dimensional image and resetting a delay quantity. SOLUTION: At first, the delay quantity of a second image as against the first image is decided, based on the motion vector of the two-dimensional image. The state of the two-dimensional image is detected, based on the motion vector of the two-dimensional image and an empharizing mode is adopted since the image is the effective one in field delay and horizontal phase control when the image is the one where two-dimensional/three-dimensional conversion by the modification of time difference (MTD) is effective (S9), the one where a subject is clearly discriminated from a background (S11) or the one where the subject is not superimposed on a screen frame (S13). The correction coefficient of the delay quantity is made to be larger than one and the correction quantity of a horizontal phase quantity is made to be larger than zero in the emphasizing mode. Thus, the delay quantity and the horizontal phase quantity are increased and a stereoscopic sense is emphasized.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for generating a three-dimensional image, and more particularly to, for example, a first method obtained from a two-dimensional image.
The present invention relates to a three-dimensional image generation method for generating a three-dimensional image using an image and a second image.

[0002]

2. Description of the Related Art In this type of prior art, as shown in FIG. 6, motion of a two-dimensional image is first detected (step S).
a), subject / background discrimination is performed based on the motion detection result (step Sb). Then, a relative motion vector between the subject and the background is calculated (Step Sc), a delay direction indicating which one of the left-eye image and the right-eye image is delayed with respect to the other, the delay amount, and the left-eye image and the right-eye image The horizontal phase amount of the image is determined (step S
d).

In such prior art, MTD (Mo
By manipulating the amount of delay of the two-dimensional image for one eye with respect to the two-dimensional image for the other eye using dified Time Difference (time difference modification), as shown in FIG. A protruding three-dimensional image is generated. Further, by using the MTD to not only manipulate the amount of delay of the two-dimensional image for the other eye with respect to the two-dimensional image for the one eye, but also manipulate the amount of horizontal phase of each image, FIG. As shown in b), not only the subject jumps out of the background, but also the subject and the background can jump out of the tube surface.

[0004]

However, conventionally, M
Regardless of whether the two-dimensional / three-dimensional conversion by TD is effective or not effective, the same control method is used, so that the generated three-dimensional video lacks sharpness in three-dimensional effect.

[0005] Therefore, a main object of the present invention is to provide a three-dimensional image generation method capable of obtaining a powerful three-dimensional image with a sharp stereoscopic effect.

[0006]

According to a first aspect of the present invention, there is provided a method for generating a three-dimensional image, comprising the steps of: generating a three-dimensional image by using a first image and a second image obtained from a two-dimensional image; Determining a delay amount of a second image with respect to a first image based on a motion vector of the two-dimensional image (a); and generating a two-dimensional image based on the motion vector of the two-dimensional image. (B) of detecting the state of (a) and determining whether or not to enter the emphasis mode for resetting the delay amount, and when the emphasis mode is set,
And (c) resetting the delay amount.

[0007] The three-dimensional image generating method according to claim 2 is
In the three-dimensional image generation method according to claim 1, when a delay amount is D and a delay amount correction coefficient is Kd (Kd> 1),
In step (c), the delay amount after resetting is calculated as D × Kd
It is to be.

[0008] According to a third aspect of the present invention, there is provided a three-dimensional image generating method.
A three-dimensional image generation method for generating a three-dimensional image using a first image and a second image obtained from a two-dimensional image,
(A) determining a horizontal phase amount of the first image and the second image based on the motion vector of the two-dimensional image; detecting a state of the two-dimensional image based on the motion vector of the two-dimensional image; Step (b) of deciding whether to set the emphasis mode for resetting, and step (c) of resetting the horizontal phase amount when the emphasis mode is set.
Is provided.

[0009] A three-dimensional image generating method according to a fourth aspect is characterized in that:
In the three-dimensional image generation method according to claim 3, assuming that the horizontal phase amount is P and the correction amount of the horizontal phase amount is Kp, the horizontal phase amount after resetting is set to P + Kp in step (c). .

[0010] According to a third aspect of the present invention, there is provided a three-dimensional image generating method.
A three-dimensional image generation method for generating a three-dimensional image using a first image and a second image obtained from a two-dimensional image,
Determining a delay amount of the second image with respect to the first image based on the motion vector of the two-dimensional image (a-1); determining a horizontal phase amount of the first image and the second image based on the motion vector of the two-dimensional image; Determining step (a-2): detecting a state of the two-dimensional image based on a motion vector of the two-dimensional image, and determining whether to set an enhancement mode for resetting a delay amount and / or a horizontal phase amount. (B)
And when the emphasis mode is set, the delay amount and / or
Alternatively, a step (c) of resetting the horizontal phase amount is provided.

[0011] A three-dimensional image generating method according to claim 6 is
6. The three-dimensional video generation method according to claim 5, wherein the delay amount is D, the horizontal phase amount is P, and the delay amount correction coefficient is Kd (Kd
> 1) Assuming that the correction amount of the horizontal phase amount is Kp, in step (c), the delay amount after resetting is set to D × Kd, and the horizontal phase amount after resetting is set to P + Kp.

[0012] According to a third aspect of the present invention, there is provided a three-dimensional image generating method.
7. The three-dimensional image generation method according to claim 1, wherein the step (b) is performed when it is detected that the two-dimensional image is an image in which the two-dimensional / three-dimensional conversion by the time difference modification is effective. The emphasis mode is set.

[0013] A three-dimensional image generating method according to claim 8 is
7. The three-dimensional image generation method according to claim 1, wherein the step (b) sets the emphasis mode when detecting that the two-dimensional image is an image in which a subject and a background can be clearly distinguished. Things.

[0014] According to a ninth aspect of the present invention, there is provided a three-dimensional image generating method,
7. The three-dimensional image generation method according to claim 1, wherein the step (b) sets the emphasis mode when detecting that the two-dimensional image is an image in which a subject does not cover a screen frame. It is.

In the three-dimensional image generation method according to the first aspect, first, a delay amount of the second image with respect to the first image is determined based on a motion vector of the two-dimensional image (step (a)). That is, if the two-dimensional image is the first image for one eye, the image obtained by delaying the two-dimensional image by a predetermined field becomes the second image. The amount of delay is determined, and the amount of projection of the subject with respect to the background is determined. Is done. Next, the state of the two-dimensional video is detected based on the motion vector of the two-dimensional video, and it is determined whether or not the emphasis mode is set (step (b)). And when the emphasis mode is set,
The delay amount is reset (step (c)).

At this time, assuming that the delay amount is D and the delay amount correction coefficient is Kd (Kd> 1), in step (c), the delay amount is reset to D × Kd. Is done. Therefore, the protrusion of the subject from the background is emphasized.

In the three-dimensional image generating method according to the third aspect, first, the horizontal phase amount of the first image and the second image is determined based on the motion vector of the two-dimensional image (step (a)). That is, the positions of the subject and the background projecting from the tube surface are adjusted. Next, the state of the two-dimensional video is detected based on the motion vector of the two-dimensional video, and it is determined whether or not the emphasis mode is set (step (b)). When the emphasis mode is set, the horizontal phase amount is reset (step (c)).

At this time, if the horizontal phase amount is represented by P and the correction coefficient of the horizontal phase amount is represented by Kp,
In step (c), the horizontal phase amount is reset to P + Kp. Therefore, the projection of the subject and the background from the tube surface is emphasized.

In the three-dimensional image generation method according to the fifth aspect, first, a delay amount of the second image with respect to the first image is determined based on a motion vector of the two-dimensional image (step (a)).
-1)) The horizontal phase amount of the first image and the second image is determined based on the motion vector of the two-dimensional image (step (a-2)). That is, the projecting amount of the subject with respect to the background is determined, and the projecting positions of the subject and the background from the tube surface are adjusted. Next, the state of the two-dimensional video is detected based on the motion vector of the two-dimensional video, and it is determined whether or not the emphasis mode is set (step (b)). When the emphasis mode is set, the delay amount and / or the horizontal phase amount is reset (step (c)).

At this time, the delay amount D, the horizontal phase amount is P, and the correction coefficient of the delay amount is Kd (K
d> 1), assuming that the correction coefficient of the horizontal phase amount is Kp, in step (c), when the delay amount D is reset, the delay amount is D × Kd, and when the horizontal phase amount P is reset. When the horizontal phase amount is reset to P + Kp and the delay amount D and the horizontal phase amount P are both reset, the delay amount is reset to D × Kd and the horizontal phase amount is reset to P + Kp.

Therefore, when the delay amount is reset, the projection of the subject from the background is emphasized, and when the horizontal phase amount is reset, the projection of the subject and the background from the tube surface is emphasized, and the delay amount and the horizontal When both phase amounts are reset, the pop-out of the subject with respect to the background is emphasized,
In addition, the projection of the subject and the background from the tube surface is emphasized.

In the step (b), when it is detected that the two-dimensional image is an image in which the two-dimensional / three-dimensional conversion by the time difference modification is effective as described in claims 7 to 9, When the two-dimensional image is detected as an image in which the subject and the background can be clearly distinguished, or
For example, when it is detected that the two-dimensional image is an image in which the subject is not over the screen frame, the two-dimensional image is set to the enhancement mode because the field delay and the horizontal phase control are effective.

[0023]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic block diagram of a main part of a three-dimensional video software conversion system 10 in which the three-dimensional video generation method of the present invention is used.

In the three-dimensional video software conversion system 10,
A normal two-dimensional video signal for displaying a two-dimensional video is input to the input terminal 12. The two-dimensional video signal input from the input terminal 12 is supplied to a video switching circuit 14 and a field memory 16.

The field memory 16 delays the input two-dimensional video signal by a predetermined field, outputs the delayed signal, and supplies it to the video switching circuit 14. The amount of delay in the field memory 16 is variably controlled by the timing control circuit 18 in units of one field.

The output side of the video switching circuit 14 is connected to an output terminal 24 for outputting a left-eye video signal L and an output terminal 26 for outputting a right-eye video signal R via phase control circuits 20 and 22, respectively. You. Video switching circuit 14
Is controlled so that the output state is switched according to the direction of movement of the subject or the background.

The phase control circuits 20 and 22 adjust the positions where the subject and the background are projected from the tube surface.
The left-eye video signal L and the right-eye video signal R are respectively subjected to horizontal phase control and output.

The two-dimensional video signal input from the input terminal 12 is supplied to a motion vector detection circuit 28.
In the motion vector detecting circuit 28, the motion between the fields,
That is, a motion vector corresponding to the moving amount (moving speed) of the subject and the background projected by the two-dimensional video signal is detected, and the detected motion vector is supplied to the CPU 30.

The motion vector is obtained by calculating the correlation between the image signal of the representative point of the previous field and the image signal of the current field in a certain detection area 34 (see FIG. 3) in the screen 32. This is a vector indicating where the representative point has moved in the current field. Such a method of detecting a motion vector is called a representative point matching method.

The CPU 30 includes a motion vector detection circuit 28
The horizontal component in the horizontal direction (hereinafter, this horizontal component is simply referred to as a “motion vector”) is extracted from the motion vectors detected in step (1), and the timing control circuit 18 is controlled accordingly.

That is, the CPU 30 first determines the subject / background based on the motion vector. In the subject / background discrimination, the maximum value and the minimum value are extracted from the motion vectors of all the detection areas 34, and the average value is obtained by calculating (maximum value + minimum value) / 2. Then, the motion vectors of all the detection areas 34 are compared with the average value. The detection area 34 in which a motion vector smaller than the average value is detected is defined as an L block, and the detection area 34 in which a motion vector larger than the average value is detected. Are defined as H blocks, respectively.

In an outer area (indicated by A in FIG. 3) in the screen 32, whether the number of H blocks or the number of L blocks is larger is determined by the central area in the screen 32 (B in FIG. 3).
), It is determined whether the number of areas of the H block or the area of the L block is larger, or the like. In the entire detection area, whether the detection area 34 of the H block is the subject and the detection area 34 of the L block is the background, L block detection area 34
Is the subject and the detection area 34 of the L block is the subject,
Is determined.

Then, the CPU 30 calculates a relative motion vector between the subject and the background.

When the movement of the subject with respect to the background is large and the relative motion vector is large, the CPU 30 controls the timing control circuit 18 so that the delay amount of the field memory 16 is reduced, and the movement of the subject with respect to the background is small or slow. When the relative motion vector is small as in the case of the motion reproduction, the timing control circuit 18 is controlled so that the delay amount of the field memory 16 increases.

Further, when the direction of the motion vector is from left to right, the CPU 30 determines that the subject from which the motion vector is detected is a subject located in front of the background, and outputs the delayed two-dimensional video signal to the right eye. If it is determined that the subject for which the motion vector is detected is a background located behind another subject, the delayed two-dimensional video signal is used as the left-eye video signal L. The switching circuit 14 is controlled. Further, when the direction of the motion vector is from right to left, the CPU 30 determines that the subject from which the motion vector is detected is a subject located in front of the background, and converts the delayed two-dimensional video signal into a left-eye video signal. L, and when it is determined that the subject for which the motion vector is detected is a background located behind another subject,
The video switching circuit 14 is controlled so that the delayed two-dimensional video signal is used as the right-eye video signal R.

The CPU 30 determines the amount of horizontal phase in order to adjust the positions of the subject and the background projecting from the tube surface, and the timing control circuit 18 transmits data of the determined horizontal phase amount to the phase control circuit 20. The phase control circuits 20 and 22 control the horizontal phase of the left-eye video signal L and the right-eye video signal R in accordance with the horizontal phase data. Phase control circuits 20 and 22
Is controlled variably by the timing control circuit 18 in units of one field.

By supplying the left-eye video signal L and the right-eye video signal R output from the output terminals 24 and 26 to a three-dimensional video display such as a lenticular lens system, the observer can enjoy a three-dimensional image. It is possible to watch a three-dimensional video having a certain depth.

The operation of the three-dimensional video software conversion system 10 will be described with reference to FIG.

First, a motion vector is detected in each detection area 34 in the screen 32 (step S1), and subject / background discrimination is performed based on the motion vector detected in each detection area 34 (step S3). Next, a relative motion vector between the subject and the background is calculated based on the motion vector detected in each detection area 34 (Step S).
5) Then, a delay direction indicating which one of the left-eye video signal L and the right-eye video signal R is delayed with respect to the other, the delay amount D, and the horizontal direction of each of the left-eye video signal L and the right-eye video signal R The phase amount P is determined (Step S7).

Next, the state of the two-dimensional image is detected.

That is, the two-dimensional image is a 2D image based on the MTD.
If the video is one in which the dimensional / three-dimensional conversion is effective, the video in which the subject can be clearly distinguished from the background, or the video does not cover the screen frame 36 (see FIG. 3) (step S9, Either of S11 and S13 is YE
S) Since the image is an image for which the field delay and the horizontal phase control are effective, the emphasis mode is set (step S15).

In the emphasis mode, the correction coefficient of the delay amount D is K
d, when the correction amount of the horizontal phase amount P is Kp, Kd = a,
Kp = b (a> 1, b> 0) is set.

Then, the delay amount becomes D.times.Kd, the horizontal phase amount becomes P + Kp (step S19), the delay amount and the horizontal phase amount are increased, the stereoscopic effect is emphasized, and the process ends.

On the other hand, the two-dimensional image is not an image in which the two-dimensional / three-dimensional conversion by the MTD is effective, an image in which the subject can be clearly distinguished from the background, and an image in which the object does not cover the screen frame 36. (Steps S9 and S1
1 and S13 are all NO), and the normal mode is set (step S17).

In the normal mode, Kd = 1 and Kp = 0 are set (step S17), and the process proceeds to step S19. Therefore, in the case of the normal mode, the delay amount D × Kd =
D, the horizontal phase amount P + Kp = P, and the delay amount D and the horizontal phase amount P obtained in step S7 are used as they are.

Here, an example of the operation for detecting the state of the two-dimensional image in steps S9 to S13 will be described with reference to FIGS. Here, (vertical)
It is assumed that 6 × (horizontal) 10 = 60 detection areas 34 are formed.

First, a case where it is determined in step S9 whether or not the two-dimensional / three-dimensional conversion based on the MTD is an effective image is described. An image in which the two-dimensional / three-dimensional conversion by the MTD is effective is, for example, an entire screen horizontal flow image or a horizontally rotated image as shown in FIG.

In this case, of the detection area 34, FIG.
The motion vector of the outer area A, the center area B, and the lower area C shown in FIG. 6 are all in the same direction, and the absolute value of the motion vector is larger than 0. Therefore, the two-dimensional image is MTD
Is determined to be an effective image.

The case where it is determined in step S11 whether or not the image is such that the subject and the background can be clearly distinguished from each other will be described. An image in which the subject and the background can be clearly distinguished is, for example, as shown in FIG. 4B, in which the subject (the car in FIG. 4B) is at the center and the background (FIG.
(Mountains and trees in (b)) are images of horizontal flow.

In this case, out of the detection area 34, FIG.
When the outer region A and (center region B + lower region C) are clearly divided into an H block and an L block, it is determined that the two-dimensional image is an image in which the subject and the background can be clearly distinguished. You.

That is, the number of H blocks included in the outer region A is equal to or more than a predetermined value β1, the number of L blocks included in (the center region B + the lower region C) is equal to or more than a predetermined value γ1, and When the difference between the number of blocks and the number of L blocks is equal to or more than a predetermined value δ1, and the difference between the number of L blocks and the number of H blocks included in (center area B + lower area C) is equal to or more than a predetermined value ε1 Or, if the number of L blocks included in the outer region A is equal to or more than the predetermined value β1, (the central region B +
The number of H blocks included in the lower area C) is equal to or more than a predetermined value γ1, the difference between the number of L blocks and the number of H blocks included in the outer area A is equal to or more than a predetermined value δ1, and (center area B + lower side) When the difference between the number of H blocks and the number of L blocks included in the area C) is equal to or larger than the predetermined value ε1, it is determined that the two-dimensional image is an image in which the subject and the background can be clearly distinguished.

Further, a case where it is determined in step S13 whether or not the subject is an image not covering the screen frame 36 will be described. The video in which the subject does not cover the screen frame 36 is, for example, a video in which the subject is located at the center of the screen as shown in FIG.

In this case, of the detection area 34, FIG.
(The outer area A + the lower area C) and the central area B are clearly divided into an H block and an L block, it is determined that the two-dimensional image is an image in which the subject does not cover the screen frame 36. You.

That is, (outer area A + lower area C)
, The number of L blocks included in the central area B is equal to or more than the predetermined value γ2, and the number of H blocks and the number of L blocks included in (outer area A + lower area C). Is greater than or equal to a predetermined value δ2 and the central region B
When the difference between the number of L blocks and the number of H blocks included in the area is equal to or more than a predetermined value ε2, or when the number of L blocks included in (the outer area A + the lower area C) is equal to or more than the predetermined value β2, the central area B Is greater than or equal to a predetermined value γ2,
The difference between the number of L blocks and the number of H blocks included in (outer area A + lower area C) is equal to or greater than a predetermined value δ2, and the difference between the number of H blocks and the number of L blocks included in central area B Is greater than or equal to the predetermined value ε2, it is determined that the two-dimensional image is an image in which the subject is not over the screen frame 36.

By operating the three-dimensional video software conversion system 10 in this manner, for example, in the normal mode, even if the subject and the background can be seen at the positions shown in FIG. , FIG. 5 (b)
The subject and the background can be seen at the positions shown in FIG.

As can be seen by comparing FIG. 5A and FIG. 5B, in the enhancement mode, by increasing the amount of horizontal phase, the entire subject and background are projected further forward. By increasing the delay amount, the protrusion of the subject from the background can be further increased.

Therefore, in the case of an image in which the field delay and the horizontal phase control are effective as described above, the pop-out amount is increased by increasing the delay amount and the horizontal phase amount more than in the normal mode. The three-dimensional effect can be emphasized. As a result, a three-dimensional image is sharpened and a powerful three-dimensional image can be obtained.

Further, an effect that a three-dimensional image which is easy on eyes and which is not tired can be provided.

In the example of the operation of the 3D video software conversion system 10 shown in FIG. 2, the delay amount and the horizontal phase amount are both enhanced in the enhancement mode.
Only one of them may be emphasized.

The present invention is not limited to the case where the delay amount D and the horizontal phase amount P are both determined to generate a three-dimensional image, but also the case where the delay amount D alone is determined to generate a three-dimensional image. , Can be applied to the case where only the horizontal phase amount P is determined to generate a three-dimensional image.

When a three-dimensional image is generated by determining only the delay amount D, steps S7, S15, S17,
The processing related to the horizontal phase amount in S19 becomes unnecessary. In this case, the protrusion of the subject from the background is emphasized.

Similarly, when a three-dimensional image is generated by determining only the horizontal phase amount P, steps S7 and S1 in FIG.
5. Processing related to delay in S17 and S19 becomes unnecessary. In this case, the projection of the subject and the background from the tube surface is emphasized.

In step S15 of FIG. 2, b> 0 in the emphasis mode. However, the present invention is not limited to this.
≤0. In this case, the subject and the background will be emphasized so as to recede from the tube surface.

Further, it goes without saying that the number of detection areas 34 formed in the screen 32 is not limited to that shown in FIG.

[0066]

According to the present invention, depending on the state of the two-dimensional image, the three-dimensional effect is enhanced more than usual, whereby a three-dimensional image with a sharp three-dimensional effect and a powerful three-dimensional image can be obtained.

[Brief description of the drawings]

FIG. 1 illustrates a 3D image generating method according to the present invention;
It is a principal part schematic block diagram which shows the three-dimensional video software conversion system.

FIG. 2 is a flowchart showing an operation of a three-dimensional image generation method.

FIG. 3 is an illustrative view showing one example of a screen;

FIG. 4 is an illustrative view showing a mode of a two-dimensional image in an emphasis mode, in which (a) is an image in which two-dimensional / three-dimensional conversion by MTD is effective, and (b) is clearly distinguishing a subject from a background; A possible image, (c), is an image in which the subject is not over the screen frame.

FIG. 5 is an illustrative view showing positions of a subject and a background in a normal mode and an emphasis mode, respectively. FIG. 5 (a) shows a case in a normal mode, and FIG. 5 (b) shows a case in an emphasis mode.

FIG. 6 is a flowchart showing a conventional technique.

FIG. 7 is an illustrative view showing a position of a subject and a background according to the related art, where (a) shows only the MTD, and (b) shows M
This is the case of TD + horizontal phase control.

[Explanation of symbols]

 Reference Signs List 10 3D video software conversion system 14 Video switching circuit 16 Field memory 18 Timing control circuit 20, 22 Phase control circuit 28 Motion vector detection circuit 30 CPU 32 Screen 34 Detection area 36 Screen frame

Claims (9)

[Claims] 1. A three-dimensional image generating method for generating a three-dimensional image using a first image and a second image obtained from a two-dimensional image, wherein the first image is generated based on a motion vector of the two-dimensional image. Determining the delay amount of the second image with respect to (a), detecting a state of the two-dimensional image based on a motion vector of the two-dimensional image, and setting an enhancement mode for resetting the delay amount A three-dimensional video generation method, comprising: a step (b) for determining whether or not the delay is set, and a step (c) for resetting the delay amount when the emphasis mode is set. 2. Assuming that the amount of delay is D and the correction coefficient of the amount of delay is Kd (Kd> 1), in the step (c), the amount of delay after reset is D ×
The three-dimensional image generation method according to claim 1, wherein Kd is set. 3. A method for generating a three-dimensional image using a first image and a second image obtained from the two-dimensional image, wherein the first image is generated based on a motion vector of the two-dimensional image. And determining the horizontal phase amount of the second image (a), detecting a state of the two-dimensional image based on a motion vector of the two-dimensional image, and setting an emphasis mode for resetting the horizontal phase amount A three-dimensional image generation method, comprising: a step (b) of determining whether or not to perform, and a step (c) of resetting the horizontal phase amount when the emphasis mode is set. 4. The method according to claim 3, wherein, if the horizontal phase amount is P and the correction amount of the horizontal phase amount is Kp, the horizontal phase amount after resetting is set to P + Kp in the step (c). 3D image generation method. 5. A method for generating a three-dimensional image using a first image and a second image obtained from a two-dimensional image, the method comprising: generating a first image based on a motion vector of the two-dimensional image. Determining the amount of delay of the second image with respect to
-1) determining a horizontal phase amount between the first image and the second image based on a motion vector of the two-dimensional image (a-2); the two-dimensional image based on a motion vector of the two-dimensional image; (B) detecting the state of the image and determining whether or not to enter an enhancement mode for resetting the delay amount and / or the horizontal phase amount; and, when the enhancement mode is set, the delay amount And / or a step (c) of resetting the horizontal phase amount. 6. The amount of delay is D, the amount of horizontal phase is P,
Assuming that the correction coefficient of the delay amount is Kd (Kd> 1) and the correction amount of the horizontal phase amount is Kp, in the step (c), the delay amount after resetting is D ×
Kd, the horizontal phase amount after resetting is set to P + Kp,
The three-dimensional image generation method according to claim 5. 7. The method according to claim 1, wherein in the step (b), when it is detected that the two-dimensional image is an image in which the two-dimensional / three-dimensional conversion by the time difference modification is effective, the emphasis mode is set. The three-dimensional image generation method according to any one of the above. 8. The three-dimensional image according to claim 1, wherein the step (b) sets the emphasis mode when detecting that the two-dimensional image is an image in which a subject and a background can be clearly distinguished. Video generation method. 9. The three-dimensional image according to claim 1, wherein the step (b) sets the emphasis mode when detecting that the two-dimensional image is an image in which a subject does not cover a screen frame. Generation method.