KR101142873B1 - Method and system for stereo image creation - Google Patents

Abstract

Disclosed are a method and apparatus for generating a stereoscopic image. The stereoscopic image generating apparatus includes a plurality of cameras for photographing a subject to generate a left image and a right image; A distance measurer having a sensor and generating distance information by measuring a distance to a subject using the sensor; A vergence controller for calculating a vergence control amount using distance information and controlling a vergence angle of the camera using the calculated vergence control amount; And an image synthesizer for comparing the left image and the right image by a predetermined scan method to detect a matching region, and clipping and combining the left image and the right image based on the detected matching region to generate a stereoscopic image.

Description

Method and apparatus for generating stereoscopic images {Method and system for stereo image creation}

The present invention relates to image processing, and more particularly, to a method and an apparatus capable of generating a multiview stereoscopic image.

Due to the two eyes separated horizontally, the parallax is generated on the retina. Due to this parallax, images of the two eyes are synthesized into one from the cerebrum to recognize the stereoscopic object.

However, images shown in photographs, videos, and TV are flat images without a three-dimensional effect because they are artificially created using video cameras or still cameras.

In order to solve this problem, a method of generating a stereoscopic image by acquiring a plurality of images of a target object using a plurality of cameras and then combining them is present. Such stereoscopic image implementation methods have a problem in that a depth is generated according to a distance from a subject, so that the effect of the stereoscopic image is lost and a lot of time and cost are required to correct it.

The present invention controls the angle of view of a plurality of cameras according to the perspective of the subject to prevent deterioration of image quality due to the occurrence of a double image, it is possible to control the angle of view in real time irrespective of the distance of the subject and time and cost for generating a stereoscopic image It is to provide a method and apparatus for generating a stereoscopic image that can significantly reduce the amount.

In addition, the present invention is to provide a method and apparatus for generating a stereoscopic image that can minimize the time consumption for generating a stereoscopic image by generating a multi-view stereoscopic image in a cascade method.

According to an aspect of the present invention, there is provided a device capable of synthesizing a stereoscopic image after obtaining a plurality of images by controlling the viewing angle of the camera according to the distance of the subject.

According to an embodiment of the present invention, a plurality of cameras for photographing a subject to generate a left image and a right image; A distance measurer having a sensor and generating distance information by measuring a distance to the subject using the sensor; A vergence controller for calculating a vergence control amount using the distance information and controlling the vergence of the camera using the calculated vergence control amount; And an image synthesizer configured to compare the left image and the right image by a predetermined scan method to detect a matching region, and to clip and synthesize the left image and the right image based on the detected matching region to generate a stereoscopic image. An apparatus for generating a stereoscopic image may be provided.

The image synthesizer performs image processing on the left image and the right image, respectively, extracts an edge to generate a left edge image and a right edge image, compares the left edge image and the right edge image by a predetermined scan method, and matches the same. An image convergence unit generating offset information about an area; And an image synthesizer configured to generate a stereoscopic image by clipping and combining the left image and the right image using the generated offset information.

The image synthesizer may further include an image converter configured to convert the left image and the right image into a black and white image, wherein the image convergence unit may generate offset information using the left image and the right image converted to the black and white image. .

The image convergence unit may detect the match region by comparing the left edge image and the right edge image based on the match region of the previous frame by a predetermined scan method.

The scan method may be any one of a spiral scan, a radial scan, and a monotonous increase or decrease scan of a left pixel or a right pixel.

The image convergence unit may classify the left edge image and the right edge image into a plurality of regions, and simultaneously compare the divided regions by the scan method to detect a matching region.

The offset information is information about a distance that the matching area is moved in the left edge image compared to the right edge image.

The vergence controller stores distances between the plurality of cameras, and the vergence control amount is further calculated by using the distances between the plurality of cameras, and the vergence control amount is used to determine the distance between the plurality of cameras and the distance information. 2 can be calculated by applying the cosine law or the arctangent function.

According to another embodiment of the present invention, a stereoscopic image generating apparatus for generating a multiview stereoscopic image, comprising: N (random natural numbers) cameras for photographing a subject to generate and output each viewpoint image; A distance measurer having a sensor for measuring a distance to the subject, and generating distance information by measuring a distance to the subject using the sensor; A vergence controller for deriving a vergence control amount using the distance information and controlling the vergence angles of the N cameras using the derived vergence control amount; A cluster unit for dividing the N viewpoint images input through the N cameras into m clusters of m pieces; K cluster convergence units for comparing matching viewpoint images in the same cluster to detect a matching region and generating cluster offset information for the detected matching region; An inter-cluster convergence unit configured to compare the last viewpoint image of the previous cluster and the first viewpoint image of the next cluster between adjacent clusters, detect a matching region, and generate inter-cluster offset information on the detected matching region; And an image synthesizer configured to generate a multiview stereoscopic image by clipping and synthesizing the N-view image using k cluster offset information and the inter-cluster offset information.

The cluster convergence unit and the inter-cluster convergence unit may include an image converter configured to convert each image into a black and white image to detect the matched region; And an image convergence unit configured to generate edge images by detecting edges in each image converted into the black and white image, and generate offset information for a matching area by comparing each edge image.

The cluster convergence unit and the inter-cluster convergence unit may generate offset information by sequentially comparing two images.

According to another aspect of the present invention, there is provided a method of synthesizing a stereoscopic image after obtaining a plurality of images by controlling the viewing angle of the camera according to the distance of the subject.

According to an embodiment of the present invention, (a) measuring the distance to the subject to generate distance information; (b) calculating a vergence control amount using the distance information, and controlling the vergence angles of the plurality of cameras provided using the calculated vergence control amount; (c) photographing a subject using a plurality of cameras having controlled viewing angles to generate left and right images, respectively; And (d) comparing the left image and the right image to detect a matched region, and generating a stereoscopic image by clipping and combining the left and right images around the detected matched region. A stereoscopic image generating method in a generating device may be provided.

The step (d) may include converting the left image and the right image into a black and white image; Generating a left edge image and a right edge image by detecting edges in the left image and the right image converted into the black and white image; Comparing the left edge image and the right edge image by a predetermined scan method to detect a matching region and generate offset information on the detected matching region; And generating a stereoscopic image by clipping and synthesizing the left and right images using the offset information.

The detecting of the matched region may include comparing the left edge image and the right edge image from the region corresponding to the offset information of the previous frame by the scanning method, if the offset information of the previous frame exists. Offset information for the corresponding current frame may be generated.

In the step (d), the detecting of the matching area may include classifying the left edge image and the right edge image into a plurality of areas, and comparing the divided areas by the scanning method simultaneously. Can be detected.

By providing a method and apparatus for generating a stereoscopic image according to an exemplary embodiment of the present invention, multi-view stereoscopic images can be generated by controlling the viewing angles of a plurality of cameras according to the perspective of a subject to prevent deterioration of image quality due to the occurrence of a double image.

In this way, by adjusting the viewing angle mechanically according to the subject's perspective, it is possible to control the viewing angle at a high speed by greatly reducing the digital image comparison area and reducing the loss due to clipping by increasing the image overlap ratio of a plurality of cameras. As a result, a high-quality stereoscopic image may be realized by an imaging device having fewer pixels.

In addition, the present invention can generate a multi-view three-dimensional image by using a cascade method to generate a multi-view three-dimensional image with a less than the number of cameras of the controller, the time for generating multi-view multi-view stereoscopic image The consumption can be minimized.

1 is a block diagram schematically illustrating a configuration of a stereoscopic image generating apparatus for generating a stereoscopic image according to a binocular viewpoint.
2 is a diagram illustrating an image scanning method.
3 is a flowchart illustrating a method in which a stereoscopic image generating apparatus synthesizes a left image and a right image to generate a stereoscopic image.
4 is a block diagram schematically illustrating an internal configuration of a stereoscopic image generating apparatus for generating a multiview stereoscopic image.
FIG. 5 is a detailed block diagram illustrating some components of the stereoscopic image generating apparatus of FIG. 4. FIG.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, it should be understood to include all transformations, equivalents, and substitutes included in the spirit and scope of the present invention. In the following description of the present invention, if it is determined that the detailed description of the related known technology may obscure the gist of the present invention, the detailed description thereof will be omitted.

The terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

 1-FIG. 2 Description

FIG. 1 is a block diagram schematically illustrating a configuration of a stereoscopic image generating apparatus for generating a stereoscopic image according to a binocular viewpoint, and FIG. 2 is a diagram illustrating an image scanning method.

Referring to FIG. 1, an apparatus for generating a stereoscopic image includes a plurality of cameras 110, a distance measurer 120, a vergence controller 130, and an image processor 140.

The first camera and the second camera are spaced apart from each other by a predetermined distance, and under the control of the stereoscopic image generating apparatus, photograph the subject to generate and output the first and second images, respectively. Hereinafter, in the present specification, for convenience of understanding and explanation, the first image generated through the first camera will be referred to as a left image, and the second image generated through the second camera will be referred to as a right image. Let's do it.

In the stereoscopic image generating apparatus, a position where the first camera and the second camera are spaced apart or a separation distance between the first camera and the second camera may be stored in advance.

The distance measurer 120 is positioned between the first camera and the second camera, has a sensor, and generates distance information by measuring a distance between a first camera and a second camera to be captured by using the sensor. To output to the gaze controller 130.

For example, the range finder 120 may be located at the center of a straight line where the first camera and the second camera are spaced apart from each other. In addition, the distance measurer 120 may include any one of an infrared sensor, an ultrasonic sensor, and a laser sensor. Since the method for measuring the distance to the subject using each sensor is obvious to those skilled in the art, a separate description thereof will be omitted. In addition, in the present specification, the sensor provided by the distance measurer 120 is specified for the purpose of convenience of understanding and explanation. Of course, in addition to the specified sensor, any device capable of measuring the distance to the subject may be applied in the same manner. .

The vergence controller 130 derives the vergence control amount by using the distance information input through the distance measurer 120, and controls and fixes the vergence angles of the first camera and the second camera according to the derived vergence control amount. It performs the function.

For example, the gaze controller 130 may include a distance between the first camera and the second camera (referred to as a camera separation distance for convenience of explanation and description) and distance information input through the distance measurer 120. The vergence control amount of the first camera and the second camera may be derived using the distance information with respect to the subject. The vergence controller 130 may derive the vergence control amount by applying the camera separation distance and the distance information to a trigonometric function (eg, the second cosine law, the arc tangent, etc.). Since this is obvious to those skilled in the art, a separate description thereof will be omitted.

After the viewing angles of the first camera and the second camera are controlled and fixed by the viewing angle controller 130, the first camera and the second camera capture the subject under the control of the stereoscopic image generating apparatus to obtain the left and right images. Each of them is generated and output to the image processor 140.

The image processor 140 analyzes the left and right images inputted through the first camera and the second camera, detects a matching region where the objects in the left image and the right image match, and matches the detected left region with the left image. And generating a stereoscopic image by clipping and synthesizing the right image.

In more detail, as illustrated in FIG. 1, the image processor 140 includes an image converter 142, an image convergence unit 144, and an image synthesizer 146.

The image converter 142 performs a function of detecting edges by outputting left and right images input through the first camera and the second camera, and outputting respective edge images corresponding to the detected edges. . Hereinafter, the edge image output from the image converter 142 corresponding to the left image will be referred to as a left edge image, and the edge image output from the right image will be referred to collectively as a right edge image. Before detecting edges of the left and right images, the image converter 142 converts the left and right images to black and white images and then performs image processing on each of the converted black and white images. Edges can be detected. The image converter 142 may detect respective edges of the left image and the right image using a Sobel mask, a Preit mask, a Roberts mask, a Laplacian, a Canny algorithm, and the like. Since the edge detection algorithm used for edge detection by the image converter 142 is obvious to those skilled in the art, a separate description thereof will be omitted. In addition, it is a matter of course that the method of detecting the edge (for example, the method using the first derivative) can be applied in the same manner.

The image convergence unit 144 detects a matching area (or coincidence point) where the objects of the left edge image and the right edge image match by using the left edge image and the right edge image input through the image converter 142. The offset information of the matched region is generated and output to the image synthesizer 146.

For example, the image convergence unit 144 sets the right edge image as the source image, sets the left edge image as the target image, and compares the target image and the source image in units of blocks of a preset size to match the edges. Detect coincidence region. Subsequently, the image convergence unit 114 may generate offset information about the detected match region. In this case, the offset information may be information on a distance, a moved pixel, or coordinate information of a matched area in the target image compared to the source image.

In addition, the image convergence unit 114 may compare the source image and the target image in units of blocks in order to detect a matching area between the source image and the target image, or may compare them in units of 1 bit plan.

FIG. 2 is a diagram illustrating a method of comparing a source image and a target image by the image convergence unit 114. Hereinafter, a method of comparing the source image and the target image in the image convergence unit 114 will be briefly described with reference to FIG. 2 for convenience of understanding and description.

The image convergence unit 114 may compare the images in the zigzag scan method with respect to the entire area of the source image and the target image. However, when comparing the source image and the target image with respect to the entire area as described above, it takes a long time to detect the matching area.

Accordingly, as illustrated in FIG. 2A, the image convergence unit 114 may detect the matched region by comparing the source image with a portion of the target image. This is because the viewing angles of the first camera and the second camera are mechanically controlled by the viewing controller 130 so that binocular disparity of the left image and the right image is reduced according to the distance from the subject. .

Basically, the overlapping region in the left image and the right image is related to the resolution of the stereoscopic image. If the viewing angle is not controlled optically, the overlapping region becomes extremely small in a close-up image in which the distance of the subject is close. On the other hand, when the mechano-optic viewing angle is controlled, the overlapping area is enlarged in the proximal region because the mesoscopic angles of the first camera and the second camera coincide with each other. Accordingly, even if the area comparing the left image and the right image is relatively smaller than the case where the mechanical optical viewing angle is not controlled, sufficiently accurate viewing angle control is possible. As in the embodiment of the present invention, the mechanical optical viewing control When the viewing angles of the first camera and the second camera are roughly controlled through and finally the viewing angle is controlled by image matching, if there is no singularity in the area comparing the source image and the target image, There is a problem that can not be controlled. Accordingly, when the singularity is found in the source image, the image convergence unit 114 searches for the singularity in the range of 10 to 20% of the effective resolution, and in the target image, the effective resolution is reflected by reflecting the offset of the screen center and the singularity of the source image. Search within 5-10% of the range. As a result, the image convergence unit 114 may search only about 5 to 10% effective planes of the X and Y planes of the entire image, thereby increasing the search efficiency by about 100 times or more.

For another example, when the viewing angles of the first camera and the second camera are controlled through the viewing controller 130, generally, a matching area of the left image and the right image is generally located near the center region. Accordingly, as shown in FIG. 2B, the image convergence unit 114 may compare the source image from the target image center region by a spiral scan to detect a matching region.

For another example, when the match region of the previous frame is stored, the image convergence unit 114 may detect the match region of the current frame by comparing the source image with the target image from the match region of the previous frame. As described above, when using the matching area of the previous frame, the time required to detect the matching area of the current frame can be reduced.

In FIG. 2 (b), the image convergence unit 114 compares the source image and the target image by the spiral scan method, but compares the source image radially from the center area of the target image or the pixel instead of the spiral scan method. Comparisons can also be made through monotonic changes in the plan.

For another example, as shown in FIG. 2C, the image convergence unit 114 divides the target image into a plurality of regions, and then compares the plurality of divided regions by a helical scan method to simultaneously compare the source images. The coincidence region with may be detected. As described above, a method of simultaneously comparing multiple areas with each other has an advantage of minimizing time for deriving a matching area.

Again, referring to FIG. 1, the image synthesizing unit 142 generates a stereoscopic image by clipping and synthesizing the left image and the right image using the offset information input through the image converging unit 144. As such, the generated stereoscopic image may be output by the image synthesizing unit 146 to an image output device (not shown).

For example, the image synthesizing unit 146 may clip the left image and the right image using the input offset information. Subsequently, the image synthesizer 146 may generate a stereoscopic image by combining the clipped left image and the right image.

[Description of Fig. 3]

3 is a flowchart illustrating a method of generating a stereoscopic image by synthesizing a left image and a right image by the stereoscopic image generating apparatus.

 Each step described below is performed by each internal component of the stereoscopic image generating apparatus, but will be collectively described as a stereoscopic image generating apparatus for the convenience of understanding and explanation.

In operation 310, the stereoscopic image generating apparatus measures distance to the subject to generate distance information. Since the method of generating the distance information by the 3D image generating apparatus is the same as described above, overlapping description will be omitted.

In operation 315, the 3D image generating apparatus derives a vergence control amount using distance information. For example, the stereoscopic image generating apparatus may derive a vergence control amount by applying distance information and camera separation distances of a plurality of cameras to a trigonometric function.

In operation 320, the stereoscopic image generating apparatus controls and fixes the viewing angle of each camera by using the derived viewing angle control amount.

In operation 325, the stereoscopic image generating apparatus controls each camera to capture (or acquire) a left image and a right image by capturing a subject.

In operation 330, the 3D image generating apparatus processes the generated left image and the right image to detect edges of the respective images to generate a left edge image and a right edge image. In this case, the 3D image generating apparatus may increase the efficiency of image processing by converting the left image and the right image into black and white images before detecting edges in the left image and the right image, respectively.

In operation 335, the 3D image generating apparatus compares the left edge image and the right edge image by using a predetermined scan method to detect a matching region and generates offset information on the detected matching region. Since this is the same as described above, overlapping description will be omitted.

In operation 340, the 3D image generating apparatus generates a 3D image by clipping and synthesizing the left image and the right image using offset information, and outputs the 3D image to an image output device (not shown).

4-5 description

4 is a block diagram schematically illustrating an internal configuration of a stereoscopic image generating apparatus for generating a multiview stereoscopic image, and FIG. 5 is a detailed block diagram illustrating some components of the stereoscopic image generating apparatus of FIG. 4.

Referring to FIG. 4, a stereoscopic image generating apparatus for generating a multiview stereoscopic image includes a camera 410, a distance measurer 420, a vergence controller 430, a cluster unit 442, a cluster convergence unit 444, and a cluster. And a liver convergence unit 446 and an image synthesizer 448.

The camera 410 includes N cameras capable of generating an image for each viewpoint by capturing an object at an N (arbitrary natural number) viewpoint. Here, when N is the square of n (any natural number), the efficiency of image processing is good. In the present specification, it is assumed that N is at most 16, but it is obvious that the number may be 16 or more depending on implementation. As illustrated at 510 of FIG. 5, the camera 410 includes N viewpoint cameras capable of capturing each viewpoint of a subject. Accordingly, the camera 410 controls N cameras to generate N-view images of the subject and output them to the cluster unit 442. Hereinafter, the image of each viewpoint will be described as a first image, a second image, ..., an N-th image for the convenience of understanding and explanation.

As described above with reference to FIG. 1, the distance measurer 420 includes a sensor, measures distance to a subject using a corresponding sensor, generates distance information, and outputs the distance information to the gaze controller 430. The distance measurer 420 may be located at the center of the center between two cameras spaced apart from each other by a predetermined distance from the N cameras (the center of gravity located in line with the two cameras) to measure the distance to the subject. In the present specification, for convenience of understanding and explanation, it will be described on the assumption that it is located at the center of the center between the first viewpoint camera and the second viewpoint camera.

For another example, the range finder 420 may be located at the center of the center of two cameras among the N viewpoint cameras.

As described above with reference to FIG. 1, the vergence controller 430 derives a vergence control amount by using the distance information input through the distance measurer 420 to control and fix the vergence of each viewpoint camera. . Since the method of deriving the vergence control amount is the same as that described with reference to FIG. 1, redundant descriptions thereof will be omitted.

In addition, the viewing angle controller 430 may control and fix the viewing angle of each viewpoint camera sequentially by the viewing angle control amount, and may simultaneously control and fix the viewing angle of each viewpoint camera using the viewing angle control amount. .

Also, the gaze controller 430 may control and fix the gaze angle of each viewpoint camera by using the angle of view control amount, and then simultaneously control the viewpoint cameras to simultaneously capture the subject and generate each viewpoint image.

The image processor 440 analyzes and synthesizes the N-view images input through the N cameras to generate a multi-view stereoscopic image.

In more detail, as illustrated in FIG. 4, the image processor 440 includes a cluster unit 442, a cluster convergence unit 444, an inter-cluster convergence unit 446, and an image synthesizer 448.

The cluster unit 442 clusters each of the N viewpoints images inputted through the multi-viewpoint image acquisition unit 410 into k (any natural number of N or less) in a predetermined method. Subsequently, the cluster unit 442 outputs the clustered image sets to the cluster convergence unit 444, respectively. In addition, the cluster unit 442 outputs the last image of each cluster and the first image of the next cluster to the inter-cluster convergence unit 446, respectively.

For example, it is assumed that each viewpoint image input through the multiview image acquisition unit 410 is a 16 viewpoint image. In this case, the cluster unit 442 classifies and clusters each view image in cluster units in k units corresponding to the number of cluster convergence units 444. For example, assume that there are four cluster convergence units 444. The cluster unit 442 may cluster the 16 viewpoint images into four bundles of three each. In the present specification, for convenience of understanding and explanation, it is assumed that the cluster unit 442 clusters the k clusters sequentially from the first viewpoint image to the Nth viewpoint image.

The cluster convergence unit 444 compares the viewpoint images in each cluster to detect a matching region, generates offset information on the detected matching region, and outputs the offset information to the image synthesis unit 448. In the present specification, for convenience of understanding and explanation, the offset information generated by the cluster convergence unit 444 will be collectively described as cluster offset information.

Referring to FIG. 5, the cluster convergence unit 444 includes an image converter 520 and an image convergence unit 530, respectively. Since the functions of the image converter 520 and the image convergence unit 530 are the same as those described with reference to FIG. 1, redundant descriptions thereof will be omitted.

In addition, as shown in FIG. 5, the cluster convergence unit 444 includes a plurality, and each cluster convergence unit 444 generates cluster offset information corresponding to each cluster to the image synthesis unit 448. Output In this case, when there are a plurality of viewpoint images in the cluster, the cluster convergence unit 444 may generate cluster offset information for a region where the viewpoint images in the cluster coincide with each other in sequence. Here, the cluster offset information may be offset information of a matching area of the plurality of viewpoint images included in the corresponding cluster. That is, cluster offset information may be accumulated about a reference point. In addition, the plurality of cluster offset information may also be accumulated and summed.

The inter-cluster convergence unit 446 analyzes adjacent inter-cluster viewpoint images to detect a matching region, generates offset information for the detected matching region, and outputs the offset information to the image synthesis unit 448. In the present specification, for convenience of understanding and explanation, the offset information generated by the inter-cluster convergence unit 446 will be collectively referred to as inter-cluster offset information.

For example, the inter-cluster convergence unit 446 may detect the matching region by analyzing the last view image of the previous cluster and the first view image of the next cluster between adjacent clusters.

The inter-cluster convergence unit 446 includes an image converter 520 and an image convergence unit 530, as illustrated at 530 of FIG. 5. Since the function of each configuration is the same as that described above, overlapping description will be omitted.

The image synthesizer 448 clips and synthesizes the N-view image using k cluster offset information input through the cluster convergence unit 444 and inter-cluster offset information input through the inter-cluster convergence unit 446. Create a multiview stereoscopic image.

That is, the image synthesizer 448 accumulates k cluster offset information and inter-cluster offset information to derive final offset information, and uses the derived final offset information to clip and synthesize an N-view image to generate a multiview stereoscopic image. Can be generated.

On the other hand, the method for generating a stereoscopic image by controlling the optical angle of the camera optically in accordance with an embodiment of the present invention is implemented in the form of program instructions that can be performed through a variety of electronic means for processing information recorded on the storage medium Can be. The storage medium may include program instructions, data files, data structures, etc. alone or in combination.

Program instructions to be recorded on the storage medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of software. Examples of storage media include magnetic media such as hard disks, floppy disks and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic-optical media such as floppy disks. hardware devices specifically configured to store and execute program instructions such as magneto-optical media and ROM, RAM, flash memory, and the like. In addition, the above-described medium may be a transmission medium such as an optical or metal wire, a waveguide, or the like including a carrier wave for transmitting a signal specifying a program command, a data structure, and the like. Examples of program instructions include not only machine code generated by a compiler, but also devices that process information electronically using an interpreter, for example, high-level language code that can be executed by a computer.

The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

Although the above has been described with reference to a preferred embodiment of the present invention, those skilled in the art to which the present invention pertains without departing from the spirit and scope of the present invention as set forth in the claims below It will be appreciated that modifications and variations can be made.

110: camera
120: distance meter
130: gaze controller
140: image processor

Claims (17)

A plurality of cameras for photographing a subject to generate a left image and a right image;
A distance measurer having a sensor and generating distance information by measuring a distance to the subject using the sensor;
A vergence controller for calculating a vergence control amount using the distance information and controlling the vergence of the camera using the calculated vergence control amount; And
An image synthesizer configured to compare the left image and the right image by a predetermined scan method, detect a matching region, and generate a stereoscopic image by clipping and combining the left image and the right image with respect to the detected matching region; Including,
The video synthesizer,
Image processing is performed on the left image and the right image to extract an edge to generate a left edge image and a right edge image, and compares the left edge image and the right edge image by a predetermined scan method to offset information on the matching area. An image convergence unit generating a; And
And an image synthesizer configured to generate a stereoscopic image by clipping and combining the left image and the right image using the generated offset information.
delete The method according to claim 1,
The video synthesizer,
The apparatus may further include an image converter configured to convert the left image and the right image into a black and white image.
And the image convergence unit generates offset information using the left image and the right image converted into the black and white image.
The method according to claim 1,
And the image convergence unit detects a match region by comparing the left edge image and the right edge image with respect to a match region of a previous frame by a predetermined scan method.
The method according to claim 1,
The scanning method may be any one of a spiral scan, a radial scan, a monotonic sensitized scan of a left pixel or a right pixel, and a monotonic sensitized scan of an upper pixel or a lower pixel.
The method according to claim 1,
And the image convergence unit divides the left edge image and the right edge image into a plurality of regions, and simultaneously compares each divided region by the scan method to detect a matching region.
The method according to claim 1,
The offset information is a stereoscopic image generating device which is information on the distance that the matching area is moved in the left edge image compared to the right edge image.
The method according to claim 1,
The vergence controller stores distances between the plurality of cameras,
And the vergence control amount is further calculated using the distance between the plurality of cameras.
The method of claim 8,
And wherein the vergence control amount is calculated by applying a distance between the plurality of cameras and the distance information to a second cosine law or an arc tangent function or using a conversion table calculated in advance.
(a) measuring distance to the subject to generate distance information;
(b) calculating a vergence control amount using the distance information, and controlling the vergence angles of the plurality of cameras provided using the calculated vergence control amount;
(c) photographing a subject using a plurality of cameras having controlled viewing angles to generate left and right images, respectively; And
(d) converting the left image and the right image into a black and white image, detecting an edge from the left image and the right image converted into the black and white image, respectively, to generate a left edge image and a right edge image, and the left edge Compare the image and the right edge image by a predetermined scan method to detect a matching region, generate offset information for the detected matching region, and clip and synthesize the left and right images using the offset information. Generating a stereoscopic image by using the stereoscopic image generating method.
delete The method of claim 10,
The detecting of the matched region may include:
If the offset information of the previous frame exists, comparing the left edge image and the right edge image from the region corresponding to the offset information of the previous frame by the scan method to generate the offset information for the current frame corresponding to the matching region A stereoscopic image generating method in a stereoscopic image generating apparatus.
The method of claim 10,
In the step (d), detecting the matching area,
And dividing the left edge image and the right edge image into a plurality of regions, and comparing the divided regions by the scanning method at the same time to detect the matched region.
In the stereoscopic image generating apparatus for generating a multi-view stereoscopic image,
N (random natural numbers) cameras for photographing a subject to generate and output respective viewpoint images;
A distance measurer having a sensor for measuring a distance to the subject, and generating distance information by measuring a distance to the subject using the sensor;
A vergence controller for deriving a vergence control amount using the distance information and controlling the vergence angles of the N cameras using the derived vergence control amount;
A cluster unit for dividing the N viewpoint images input through the N cameras into m clusters of m pieces;
K cluster convergence units for comparing matching viewpoint images in the same cluster to detect a matching region and generating cluster offset information for the detected matching region;
An inter-cluster convergence unit configured to compare the last viewpoint image of the previous cluster and the first viewpoint image of the next cluster between adjacent clusters, detect a matching region, and generate inter-cluster offset information on the detected matching region; And
and an image synthesizer configured to generate a multiview stereoscopic image by clipping and synthesizing the N viewpoint images using k cluster offset information and the inter-cluster offset information.
The method of claim 14,
The cluster convergence unit and the inter-cluster convergence unit,
An image converter for converting each image into a black and white image to detect the matching region; And
And an image convergence unit configured to generate edge images by detecting edges in each image converted into the black and white image, and to generate offset information for a matching area by comparing each edge image.
The method of claim 14,
The cluster convergence unit and the inter-cluster convergence unit,
A stereoscopic image generating device generating offset information by sequentially comparing two images.
The method of claim 14,
The vergence control unit calculates the vergence control amount by further using the position information of the camera,
And calculating the vergence control amount by applying the distance information and the position information to a conversion table using a trigonometric function or a pre-calculated trigonometric function.

Images