JP3234412B2 - Motion vector detection device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motion vector detecting device suitable for use in a three-dimensional image conversion device for converting a two-dimensional image into a three-dimensional image having parallax.

[0002]

2. Description of the Related Art Conventionally, to obtain a three-dimensional image having parallax between left and right, a two-channel stereoscopic image signal (three-dimensional image signal) obtained by imaging with a dedicated stereoscopic imaging device is recorded on a stereoscopic VTR or the like. It was necessary to reproduce this and reproduce it on a dedicated three-dimensional display or the like.

Therefore, according to this method, existing two-dimensional image software cannot be used, and new three-dimensional image software has to be produced, which causes an increase in the cost of a three-dimensional image reproduction system. .

[0004] For this reason, the present applicant has previously filed Japanese Patent Application No.
In -10583, a two-dimensional image is delayed in accordance with the motion of the image. For example, by using the original signal as a right-eye image and the delayed image as a left-eye image, for a horizontally moving image, A method for obtaining a three-dimensional image having parallax has been proposed.

First, the principle of this method will be described.
In a video scene in which the background does not change and the subject moves from left to right as shown in FIG. 4A, when a fixed time difference is provided between the reproduced right-eye video and left-eye video as shown in FIG.
The position is different by the movement of the subject, which becomes parallax as shown in FIG. 4B and 4C.
Indicates the field number.

Next, a three-dimensional image converter based on this principle will be described with reference to FIG. The input terminal 1 receives a two-dimensional video signal. One of the two-dimensional video signals is supplied to the video switching circuit 2.

The other of the two-dimensional video signals is supplied to a field memory 5. The field memory 5 is variably controlled by the memory control circuit 6 in a field unit from a delay amount of 0 to a maximum of 60 fields (about 1 second in the NTSC system).

The output of the field memory is supplied to the video switching circuit 2. This video switching circuit 2
The outputs are connected to an output terminal 3 for outputting a left-eye video signal L and an output terminal 4 for outputting a right-eye video signal R, respectively, and are controlled so that the output state is switched according to the direction of movement of the subject.

The other one of the two-dimensional video signals is supplied to a motion vector detecting circuit 7, and is supplied to a CPU 8 after a motion vector corresponding to a motion between fields is detected.
The CPU 8 extracts a horizontal component from the motion vector and controls the memory control circuit 6 accordingly. That is, when the motion of the subject is large and the motion vector is large, the delay amount of the field memory 5 is controlled to be small. When the motion of the subject is small or the motion vector is small as in slow motion reproduction, the delay amount is reduced. It is controlled to increase.

Further, the CPU 8 controls the video switching circuit so that the two-dimensional video signal is used as the left-eye video signal when the direction of the motion vector is from left to right, and the delayed two-dimensional video signal is used as the left-eye video signal when the motion vector is reversed. 2 is controlled.

Therefore, in a scene in which the subject moves in the horizontal direction in the two-dimensional video signal, a parallax corresponding to the speed of movement is generated. The output terminals 3 and 4
If the left and right video signals are supplied to a lenticular type display without glasses as described in Japanese Patent Application Laid-Open No. 3-65943, for example, a stereoscopic image having a partially three-dimensional appearance is pseudo even if it is a two-dimensional video signal. Can be reproduced.

For the motion vector detection required for the three-dimensional image converter, a representative point matching method used in a camera shake correction device of a video camera can be used. For this representative point matching method, see National Tec
hnical Report Vol.37 No.3Jun.1991, pp.48-54, and will be described below.

[0013] This representative point matching method refers to one field (or one frame) at a plurality of fixed representative points.
The previous video signal level is compared with the video signal level of the sampling point in the detection area including the representative point, and the sampling point of the current field having the smallest correlation and the highest correlation is obtained. The position difference (shift) is specified as the motion of the subject, that is, as a motion vector.

Next, the representative point matching method will be specifically described. FIG. 6 shows a valid screen, in which 1 is displayed.
Two detection blocks are set. Further, each detection block is divided into 15 detection areas. As shown in an enlarged manner in FIG. 7, there are a plurality of sampling points in each detection area, and one of the sampling points at the center is set as a representative point.

FIG. 8 shows a block diagram of a motion vector detecting device based on the representative point matching method. The digital video signal input to the input terminal 8 is supplied to the representative point memory 9 and the correlation value calculation circuit 10. The representative point memory 9 stores digital data corresponding to the luminance level of each representative point.
In the correlation value calculation circuit 10, the absolute value of the difference between the digital data corresponding to the luminance level of the video signal of the current frame and the digital data one frame (or one field) before from the representative point memory 9, that is, the representative point Then, a correlation value between the luminance and the sampling point shifted from the representative point in the detection area is calculated. This correlation value is cumulatively added to each sampling point having the same shift with respect to the representative point in the detection block in the cumulative addition circuit 11 including the correlation value memory and the adder. This cumulative addition is performed for each detection block.

The output of the accumulation circuit is supplied to a minimum value detection circuit 12 and an average value calculation circuit 13. Minimum value detection circuit 1
2 detects the position of the sampling point where the correlation accumulated value is the minimum and the minimum value thereof. The average value calculation circuit 13 calculates the average of the accumulated correlation values. And the determined minimum position,
The minimum value and the average value are supplied to the microcomputer 14. The microcomputer 14 first extracts a motion vector based on the minimum position for each detection block by the software processing. Next, among the motion vectors, those having a minimum value / average value smaller than a predetermined threshold value are removed as having low reliability, and one motion vector is specified from the remaining motion vectors. The address, timing, and the like of the representative point memory 9, the cumulative addition circuit 11, and the average value calculation circuit 13 are controlled by the control circuit 15, and the position of the representative point and the detection area are set.

When the above-described motion vector detecting device is used in a three-dimensional image converting device, the motion vector detecting device is used over the entire effective screen.
Since it is necessary to detect a motion vector, a large number of detection blocks are arranged over the entire effective screen.

However, in video software such as a movie, there is a horizontally long so-called letterbox video having an aspect ratio different from that of the NTSC system as shown in FIG. In this letterbox image, a black level area B where no image exists above and below appears. For this reason, when the black level area B and the detection block overlap, there is a drawback that an accurate motion vector cannot be detected from the detection block.

[0019]

SUMMARY OF THE INVENTION The present invention has been made in view of the above points, and provides a motion vector detecting apparatus capable of accurately detecting a motion vector even when input video software is letterbox video. Is what you do.

[0020]

According to the present invention, a plurality of detection areas are provided on a screen, and a correlation value calculating means for calculating a correlation value at a predetermined transition for each of the detection areas, and In a motion vector detecting device including a means for obtaining a motion vector for each of the detection areas and a motion vector specifying means for specifying a motion vector of the entire screen from the motion vector, a plurality of images having different aspect ratios are provided.
A video discriminating means for discriminating an image, and
Detection region changing unit and the motion vector detection apparatus der formed by providing a for changing the detection area Te is, specifically, the video determination
Means it is discriminated black level region of the input video signal.

Further, the picture discriminating means discriminates the black level area based on the luminance level of the picture signal. Further, the video discriminating means discriminates a black level area based on the reliability of the correlation value in each of the detection areas.

[0022]

According to the present invention, the picture discriminating means discriminates a boundary between a black level area and a picture area when a letterbox picture is input. Based on the output of the determination means, the detection area changing means changes the position of the detection area so as not to overlap the black level area.

The detection area changing means invalidates the motion vector detected from the detection area overlapping the black level area, and does not use the motion vector for specifying the motion vector of the entire screen.

[0024]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic block diagram of a motion vector detection circuit according to the present embodiment. The same parts as those in FIG.

A feature of this embodiment is that a video discriminating circuit 16 for discriminating the level of an input video signal and outputting a vertical address indicating a settable range of a detection block is provided. The point is that the position setting is automatically changed.

That is, at the time of normal video input, the detection block is set over the entire screen, but at the time of letterbox video input, if the setting is left as it is, the detection block and the black level area are set as shown in FIG. Will overlap. At this time, the video determining circuit 16 determines the boundary between the black level area and the video area, and outputs the boundary as a vertical address.
With this vertical address, the microcomputer can operate as shown in FIG.
The setting of the position is changed so that the detection block is concentrated at the center as shown in FIG. Next, when the input image returns to normal, the detection block spreads up and down again.

Next, the details of the video discrimination circuit 16 will be described with reference to FIG. First, an input video signal is supplied to a video detection circuit 161.
And outputs a high when the input is higher than the reference level, and outputs a low when the input is lower than the reference level. As a result, when a letterbox image is input, black level areas at the top and bottom of the screen are detected. That is, a low detection output is output in the upper and lower black level areas, and a high detection output is output in the video signal area. This detection output is supplied to the video start vertical address detection circuit 162 and the video end vertical address detection circuit 163, respectively.
Supplied to The video start vertical address detection circuit 162 detects the vertical address when the input changes from low to high, that is, when the input changes from the black level area at the top of the screen to the video area. The video end vertical address detection circuit 163 detects a vertical address when the input switches from high to low, that is, when the video area changes to a black level area at the bottom of the screen. Then, the respective vertical addresses are supplied to the microcomputer 14.

On the other hand, in the case of a normal input video having no black level area, the video start vertical address detection circuit 162 detects the vertical address when the vertical blanking period changes to the video area, and detects the video end vertical address. The circuit 163 detects a vertical address at the time of changing from a video area to a vertical blanking period.

As described above, the microcomputer 1
In the case of letter box video input, the distribution of the detection blocks is concentrated in the vertical direction so that the detection blocks do not overlap with the black level regions as shown in FIG. Change the settings. That is, the microcomputer 14 outputs data indicating the position of the detection block and data indicating the position of the representative point to the control circuit 15, and sets the detection block position and the representative point position.

In the above-described embodiment, the detection block is moved during the letterbox video. However, the motion vector detected in the detection block overlapping the black level area is invalidated, and the motion vector of the entire screen is invalidated. It is possible not to use it for specific purposes.

In addition to the method based on the video signal level as in the above-described embodiment, letterbox video is distinguished from the correlation value of a series of blocks at the upper end or lower end of one row of detection blocks arranged in the horizontal direction. By determining the reliability, it may be determined whether or not the detected block is in the black level area.

[0032]

As described above, according to the present invention, even if the input video software is a letterbox video, it is detected and automatically changed from the black level area by changing the representative point and the detection block. It is possible to prevent the detected motion vector from being erroneously adopted, and to accurately detect the motion vector of the entire screen.

[Brief description of the drawings]

FIG. 1 is a block diagram of a motion vector detection device according to an embodiment of the present invention.

FIG. 2 is a block diagram of a video discrimination circuit according to the embodiment.

FIG. 3 is an explanatory diagram of a relationship between a detection block and a black level area in the embodiment.

FIG. 4 is a principle diagram of three-dimensional image conversion.

FIG. 5 is a schematic block diagram of a three-dimensional image conversion device.

FIG. 6 is an explanatory diagram of a detection block in the representative point matching method.

FIG. 7 is an explanatory diagram of a detection area in the representative point matching method.

FIG. 8 is a block diagram of a conventional motion vector detection device.

FIG. 9 is an explanatory diagram of a conventional relationship between a detection block and a black level area.

[Explanation of symbols]

 9 Representative point memory 10 Correlation value calculation circuit 11 Cumulative addition circuit 12 Minimum value detection circuit 13 Average value calculation circuit 14 Microcomputer (detection area changing means) 15 Control circuit 16 Video discrimination circuit

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Yukio Mori 2-5-5-1 Keihanhondori, Moriguchi-shi, Osaka Sanyo Electric Co., Ltd. (56) References JP-A-4-58687 (JP, A) JP JP-A-5-130484 (JP, A) JP-A-5-219420 (JP, A) JP-A-6-181563 (JP, A) JP-A-4-271675 (JP, A) (58) Fields investigated (Int) .Cl. ⁷ , DB name) H04N 13/00-15/00

Claims (4)

(57) [Claims] A plurality of detection areas are provided on a screen, a correlation value calculating means for calculating a correlation value in a predetermined transition for each of the detection areas, and a motion vector for each of the detection areas based on the correlation value. means for determining, in the motion vector detecting apparatus comprising a motion vector specifying means for specifying a motion vector of the entire screen from the motion vector, the image discrimination hand for determining a plurality of images whose aspect ratios are different
A motion vector detection device comprising: a stage; and a detection area changing means for changing the detection area based on an output of the means. 2. A method according to claim 1, wherein a plurality of detection areas are provided on the screen, a correlation value calculating means for calculating a correlation value in a predetermined transition for each of the detection areas, and a motion vector for each of the detection areas based on the correlation value. , And a motion vector specifying device for specifying a motion vector of the entire screen from each of the motion vectors. A video discriminating device for discriminating a black level region of an input video signal, and an output of the device. And a detection area changing means for changing the detection area based on the motion vector. 3. The motion vector detecting device according to claim 2, wherein said video discriminating means discriminates a black level area based on a luminance level of a video signal. 4. The motion vector detecting device according to claim 2, wherein said video discriminating means discriminates a black level area based on reliability of a correlation value in each of said detection areas.