WO2011114683A1 - 立体視用画像位置合わせ装置、立体視用画像位置合わせ方法、及びそのプログラム - Google Patents
立体視用画像位置合わせ装置、立体視用画像位置合わせ方法、及びそのプログラム Download PDFInfo
- Publication number
- WO2011114683A1 WO2011114683A1 PCT/JP2011/001455 JP2011001455W WO2011114683A1 WO 2011114683 A1 WO2011114683 A1 WO 2011114683A1 JP 2011001455 W JP2011001455 W JP 2011001455W WO 2011114683 A1 WO2011114683 A1 WO 2011114683A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- point
- pair
- stereoscopic
- matrix
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G03—PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
- G03B—APPARATUS OR ARRANGEMENTS FOR TAKING PHOTOGRAPHS OR FOR PROJECTING OR VIEWING THEM; APPARATUS OR ARRANGEMENTS EMPLOYING ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ACCESSORIES THEREFOR
- G03B35/00—Stereoscopic photography
- G03B35/08—Stereoscopic photography by simultaneous recording
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
- H04N13/106—Processing image signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
- H04N13/106—Processing image signals
- H04N13/122—Improving the 3D impression of stereoscopic images by modifying image signal contents, e.g. by filtering or adding monoscopic depth cues
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
- H04N13/106—Processing image signals
- H04N13/128—Adjusting depth or disparity
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/20—Image signal generators
- H04N13/204—Image signal generators using stereoscopic image cameras
- H04N13/207—Image signal generators using stereoscopic image cameras using a single 2D image sensor
- H04N13/221—Image signal generators using stereoscopic image cameras using a single 2D image sensor using the relative movement between cameras and objects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/20—Image signal generators
- H04N13/204—Image signal generators using stereoscopic image cameras
- H04N13/239—Image signal generators using stereoscopic image cameras using two 2D image sensors having a relative position equal to or related to the interocular distance
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/20—Image signal generators
- H04N13/261—Image signal generators with monoscopic-to-stereoscopic image conversion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/30—Image reproducers
- H04N13/366—Image reproducers using viewer tracking
- H04N13/373—Image reproducers using viewer tracking for tracking forward-backward translational head movements, i.e. longitudinal movements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N2013/0074—Stereoscopic image analysis
Definitions
- the present invention relates to a stereoscopic image alignment device that detects corresponding points from a plurality of image pairs and aligns them so as to satisfy a certain standard, and in particular, stereoscopic viewing is possible in a stereoscopic display device.
- the present invention relates to an image alignment apparatus for stereoscopic viewing and the like that aligns image pairs as described above.
- Stereo imaging has been studied for many years in this technology. Stereoscopic images allow viewers to feel depth and feel a sense of realism, and in the home entertainment industry, it is now a hot topic. Stereo imaging applications range from military navigation to medical imaging, industrial testing, consumer electronics, and more.
- the principle of stereoscopic imaging and display is well known. That is, the left-eye and right-eye images constituting the image pair are taken from slightly different viewpoints (usually viewpoints whose positions are shifted in the horizontal direction). By displaying the left eye image separately for the left eye and the right eye image for the right eye separately, the human brain combines them and perceives these parallaxes as depth.
- this image pair can be obtained by using a photographing device (for example, a digital stereo camera) having two image sensors.
- a photographing device for example, a digital stereo camera
- two image sensors including the optical system
- the captured image may be distorted due to lens distortion or camera structure.
- a difference occurs between the horizontal direction and the vertical direction.
- the vertical and horizontal differences between two images are commonly known as parallax.
- at least the vertical parallax must be eliminated and the horizontal parallax must be adjusted appropriately. is there.
- an image pair for stereoscopic viewing can be obtained by capturing one image and moving the camera to capture the next image.
- a monocular sensor device There are two advantages of using a monocular sensor device. One is that the cost is low compared to the case of using a twin-lens (stereo) sensor device. Another advantage is the flexibility to change the distance between the left image and the right image depending on the scene content.
- the main problem with this method is that when a handheld DSC (Digital still camera) is used, the left image and the right image are misaligned mainly due to camera shake.
- DSC Digital still camera
- Patent Document 1 discloses a method and apparatus for correcting an optical misalignment of a stereoscopic image capturing system such as a stereoscopic electronic endoscope. Is disclosed. In this method, calibration information is determined by calculating a conversion rate, a rotation rate, and a scale using a calibration target. Then, alignment of the image pair is performed using the determined calibration information.
- Patent Document 2 Japanese Patent Laid-Open No. 2003-244500 discloses an imaging system that captures right and left parallax images in time series to obtain a stereoscopic image of an object.
- the LCD viewfinder is used to guide the user to align the two images during shooting.
- Patent Document 3 discloses a method for determining a parallel transformation of a pair of images.
- US Pat. No. 6,606,404 discloses a system and method for calculating a projective transformation correction for stereoscopic processing of a three-dimensional object. Yes.
- An object of the present invention is to convert the left and right images so that the epipole is mapped to infinity and the epipolar line is parallel to the horizontal axis. By this method, an image suitable for stereoscopic viewing can be output.
- Patent Document 1 cannot be applied when the position of the camera is changed due to, for example, an impact. Also, this method is difficult to apply to image pairs taken with a monocular device. Also, this method is difficult to apply to image pairs taken with a variable baseline stereo camera.
- Patent Document 3 is a method particularly configured for a stereo camera in which the distance between two sensors is fixed and the left and right images are taken simultaneously. Therefore, it is difficult to apply this method to image pairs photographed with a monocular device. Also, this method is difficult to apply to image pairs taken with a variable baseline stereo camera.
- the present invention does not depend on camera parameters and can also be applied to image pairs taken with a monocular camera and a variable baseline camera, and automatically aligns image pairs for stereoscopic viewing in a shorter time than before.
- An object of the present invention is to provide a stereoscopic image alignment device and the like.
- a stereoscopic image registration device acquires an image pair including a left-eye image and a right-eye image corresponding to the left-eye image.
- a corresponding point detection unit that detects a corresponding point that is a set with a certain second point, and the first parallax is such that the vertical parallax between the first point and the second point is minimized and the epipolar constraint is satisfied.
- a first matrix calculating unit that calculates a projective transformation matrix that transforms the points, a converting unit that converts the first image using the projective transformation matrix, and a third image that is the converted first image.
- An output unit configured to output an image and the second image;
- the stereoscopic image alignment apparatus can perform alignment so as to eliminate the vertical parallax of the left and right images included in the image pair.
- the parallax necessary for stereoscopic viewing is horizontal parallax, and the vertical parallax hinders the natural stereoscopic effect.
- the stereoscopic image according to this embodiment is used.
- the alignment device is effective.
- the calculation method of the projective transformation matrix in the present embodiment can be automated by a computer, and does not depend on camera parameters, and can also be applied to image pairs taken with a monocular camera and a variable baseline camera.
- the two images are converted into two images as in the prior art. Roughly half the matrix calculation is required rather than processing. Therefore, the calculation can be performed in a shorter time than in the past.
- the image pair acquisition unit displays two images simultaneously captured by two image sensors and two lenses, one image sensor and two lenses, or two image sensors and one lens. You may acquire as a pair.
- the image pair acquisition unit may acquire two images taken at different times from different viewpoints as the image pair using one image sensor and one lens.
- a second matrix for calculating an oblique transformation matrix for transforming the third image so that the coordinate system of the third image has orthogonality and has the same aspect ratio as the first image A calculation unit; and a matrix synthesis unit that generates a composite transformation matrix by combining the projection transformation matrix and the oblique transformation matrix, and the transformation unit uses the synthetic transformation matrix to generate the first image. May be converted.
- the second matrix calculation unit has the orthogonality of the coordinate system of the inclined third image (that is, the horizontal axis and the vertical axis in the image plane are vertical), and the second matrix calculation unit An oblique transformation matrix for transforming the third image is calculated so as to have the same aspect ratio as that of the first image.
- the acquired image pair can be output as an image pair capable of more accurate stereoscopic viewing.
- the image processing apparatus further includes a viewing zone calculation unit that calculates a viewing zone that is a region common to the first image and the second image, and the second matrix calculation unit includes the viewing zone in the third image.
- the oblique transformation matrix may be calculated so that the coordinate system of the portion included in the image has orthogonality and has the same aspect ratio as the portion included in the viewing area in the first image. .
- the image area that hinders the effect of stereoscopic vision can be deleted in advance by the viewing area calculation unit.
- the viewing area calculation unit it is possible to perform processing only on an image area effective for comfortable stereoscopic viewing, thereby shortening the processing time and outputting an image pair capable of more comfortable stereoscopic viewing.
- the corresponding point detection unit includes a plurality of small regions uniformly distributed within a predetermined range from the center of each of the first image and the second image included in the image pair. It is also possible to select and detect the corresponding point from the small area.
- the corresponding point detection unit can detect the target point by paying attention to the region where the corresponding point is easy to find among the regions included in the target image. As a result, the detection accuracy of the corresponding points is improved, and the image correction processing using the more accurate projective transformation matrix becomes possible.
- a blur region detection unit that detects a blur region in the image pair may be further provided, and the corresponding point detection unit may detect the corresponding point from a region other than the blur region in the image pair.
- the output unit obtains adjustment information indicating at least one of an input operation from the viewing user, a size of the display device, and a distance between the display device and the viewing user, and based on the adjustment information,
- the parallax, the 3D pop-up, and the deep-in effect of the third image and the second image may be adjusted and output.
- the output unit can give the optimal stereoscopic effect to the output image based on the index of comfort and safety based on the context of the place.
- the image pair acquisition unit acquires image pair candidates that are candidates for a plurality of image pairs, and (B) each of the plurality of image pair candidates is common to each of the other image pair candidates.
- a common viewing area that is the viewing area is specified, quality information that is information on at least one of the light amount, the luminance, the color, and the sharpness in the common viewing area is acquired, and (C) indicated by the quality information
- the image pair candidate having the best image quality may be selected as the image pair.
- the image pair acquisition unit selects an image from which a higher quality stereoscopic effect can be obtained, Can be paired.
- the image pair acquisition unit may (1) compress the image group including the left-eye image and the right-eye image and store them in the memory card, or (2) compress the image group. If the image group is transmitted from an external device via a network, the compressed image group may be decoded.
- the image pair acquisition unit can make corrections that enable more appropriate stereoscopic viewing of an image acquired via an external storage medium or a communication network.
- the corresponding point detection unit may calculate a rotation parameter and a shift parameter used for creating the projective transformation matrix so as to coincide with an output from a gyro sensor that outputs information corresponding to a rotation angle.
- the corresponding point detection unit can calculate the corresponding point faster and more accurately by using the corresponding point detection algorithm related to the related technology and the actually measured value by the gyro sensor.
- the present invention can be realized not only as such a stereoscopic image alignment apparatus, but also as a stereoscopic image alignment method using characteristic means included in the stereoscopic image alignment apparatus as a step. Or, it can be realized as a program for causing a computer to execute such characteristic steps. Such a program can be distributed via a recording medium such as a CD-ROM (Compact Disc Only Memory) and a transmission medium such as the Internet.
- a recording medium such as a CD-ROM (Compact Disc Only Memory)
- a transmission medium such as the Internet.
- the present invention can be realized as a semiconductor integrated circuit (LSI) that realizes part or all of the functions of such a stereoscopic image alignment device, or a stereoscopic image including such a stereoscopic image alignment device. It can be realized as a visual image alignment system.
- LSI semiconductor integrated circuit
- Stereoscopic images that are independent of camera parameters and can be applied to image pairs taken with a monocular camera and variable baseline camera, and automatically align image pairs for stereoscopic viewing in a shorter time than before.
- An alignment device or the like can be provided.
- the present invention can be processed automatically, it has an effect over the prior art in Patent Document 2. Moreover, since this invention does not depend on a calibration target, it has the effect over the prior art in patent document 1.
- FIG. in addition, the present invention is designed for a monocular camera and can be applied to a pair of images taken by a variable baseline twin-lens camera, and thus exceeds the prior art in Patent Document 1 and Patent Document 3. Has an effect.
- FIGS. 5a, 5b, and 5c The method in U.S. Patent No. 6,057,051 shows the camera and image orientation prior to alignment, shown as left image 402 and right image 404 in Fig. 5a, as parallelized left image 406 and parallelized right image 408 in Fig. 5b.
- the purpose is to realize alignment (parallelization) in such a direction.
- the epipole is mapped to infinity, and the epipolar line is parallel to the horizontal axis.
- the present invention is not intended to achieve this. Instead, the present invention aims to achieve the alignment shown in FIG. 5c.
- the right image 404 is transformed (deformed) as indicated by the right image 410.
- the present invention performs image conversion of either one of the left image and the right image. Therefore, the present invention has an effect that the processing time can be shortened because the complexity is lower than the prior art in Patent Document 4 (processing is simple).
- FIG. 1 is a diagram illustrating a positional deviation of a pair of stereoscopic images captured using a hand-held monocular imaging device.
- FIG. 2 is a block diagram showing an overview of the stereoscopic image alignment apparatus according to Embodiment 1 of the present invention.
- FIG. 3 is a block diagram showing a detailed functional configuration of the image alignment module according to Embodiments 1 and 2 of the present invention.
- FIG. 4a is a flowchart showing processing of the image alignment module according to Embodiments 1 and 2 of the present invention.
- FIG. 4b is a flowchart showing the processing of the image alignment module including the blur region operation function according to Embodiments 1 and 2 of the present invention.
- FIG. 5a is a diagram showing the direction of the camera and the image before alignment.
- FIG. 5b is a diagram showing the direction of the camera and image after parallelization according to the prior art.
- FIG. 5c is a diagram illustrating the direction of the camera and the image after alignment by the stereoscopic image alignment apparatus according to Embodiments 1 and 2 of the present invention.
- FIG. 6a is a diagram illustrating the concept of a viewing area or an overlapping area from an image pair.
- FIG. 6B is a diagram illustrating the aspect ratio of the viewing area or the overlapping area before and after performing the transformation based on the concept of orthogonality and the projective transformation matrix.
- FIG. 7 is a block diagram of an imaging apparatus according to Embodiment 2 of the present invention.
- FIG. 8 shows one of the results showing the effects of the stereoscopic image alignment device according to the first and second embodiments of the present invention.
- FIG. 9 shows an application example in which the present invention is used as a 3D processing engine.
- FIG. 10A is a diagram illustrating a case where stereoscopic display is performed with normal parallax.
- FIG. 10B is a diagram illustrating a case where stereoscopic display is performed with zero parallax.
- FIG. 10A is a diagram illustrating a case where stereoscopic display is performed with normal parallax.
- FIG. 10B is a diagram illustrating a case where stereoscopic display is performed with zero parallax.
- FIG. 10c is a diagram illustrating a case where stereoscopic display is performed with negative parallax.
- FIG. 11 is a diagram illustrating user interaction communicating with a 3D processing engine in a 3D display via a 3D display button.
- FIG. 12a is a first diagram illustrating feature point extraction and image region selection for generating a deformation matrix.
- FIG. 12b is a second diagram illustrating feature point extraction and image region selection for generating a deformation matrix.
- FIG. 13 is a diagram for explaining an example of a method of calculating the oblique transformation matrix in the first and second embodiments of the present invention.
- FIG. 14 is a conceptual diagram showing an image of the image conversion process using the projective transformation matrix and the oblique transformation matrix in the first and second embodiments of the present invention.
- FIG. 1 shows an example in which an image pair is photographed using a monocular imaging device (that is, provided with one lens and one image sensor).
- the left image 102 and the right image 104 are continuously photographed from different viewpoints using the monocular digital camera 106.
- the moving distance of the camera between the shooting of the left image 102 and the shooting of the right image 104 is indicated by a distance 108.
- the user captures the left image 102 at a certain position and then moves the hand to the right by the distance 108 to capture the right image 104.
- the captured images (left image 102 and right image 104) are not aligned and include the effects of transformation and rotation due to camera movement (eg, by hand).
- the distance 108 may vary depending on the scene. For example, the distance 108 becomes shorter when the object is near, and becomes longer when the object is far.
- the left image 102 and the right image 104 photographed in this example can be appropriately aligned by the stereoscopic image alignment device according to the present invention.
- FIG. 2 is a block diagram of a stereoscopic image alignment apparatus 200 according to an embodiment of the present invention.
- the stereoscopic image registration apparatus 200 includes an image pair acquisition unit 205, an image registration module 208, an output unit 210, and an internal buffer 212.
- the image pair acquisition unit 205 acquires an image pair including a left eye image and a right eye image corresponding to the left eye image.
- the image pair acquisition unit 205 may acquire an image pair from a photographing device (not shown) including an image sensor and a lens, or may acquire an image pair via a storage medium or a network.
- the image pair acquisition unit 205 acquires an image pair from the imaging device, (1) two image sensors and two lenses, (2) one image sensor and two lenses, or (3) two It is conceivable to acquire two images captured at the same time as an image pair by an imaging device including either an image sensor or one lens.
- the image pair acquisition unit 205 may acquire two images captured at different times from different viewpoints as an image pair using one image sensor and one lens.
- the image pair acquisition unit 205 includes a storage medium reader 202 and an image decoder 204.
- the storage medium reader 202 acquires image data from the storage medium.
- the image decoder 204 decodes the image data as necessary.
- the image decoder 204 may decode encoded data acquired via a network, for example. That is, the image pair acquisition unit 205 (1) when the image group including the left eye image and the right eye image is compressed and stored in the memory card, or (2) the image group is compressed, In addition, when the image is transmitted from the external device via the network, the compressed image group may be decoded.
- the image alignment module 208 aligns the uncompressed image pair S23 by performing image conversion. Intermediate data associated with the alignment process is recorded in the internal buffer 212. The aligned image pair S25 is transmitted to the output unit 210.
- the output unit 210 outputs an image pair including an image converted by alignment. Specifically, the output unit 210 adjusts the output image format according to a preferred output type. For example, the output unit 210 may adjust the image format to match the display format that can be displayed by the display device 214 in order to display on the display device 214.
- Such formats include formats for autostereoscopic devices that are viewed without glasses, gray / color anaglyphs (viewed using glasses), interlace formats, checkerboard formats, or active shutter glasses. Other formats for a frame sequential stereoscopic display device may be included.
- the registered image pair S ⁇ b> 26 may be stored / transmitted using the storage / transmission device 216.
- Examples of storage devices include, but are not limited to, flash-based memory cards, hard drives, and optical drives.
- Examples of the transmission device include, but are not limited to, an HDMI interface, a USB interface, a wireless interface, and a direct-to-printer interface. Data processed by the storage device or the transmission device may be arbitrarily reversible or lossy compressed.
- the aligned image pair S26 can also be output for printing by the printer 218.
- the printer 218 may include a lenticular sheet that builds an illusion of depth on the printed image.
- the image stored in the storage / transmission device 216 may be transmitted directly to the printer 218 or may be transmitted to the display device 214 via the signal S27. If necessary, a feedback signal S29 that may include the width of the display may be fed back to the image registration module 208.
- the image alignment module 208 and other internal modules included in the stereoscopic image alignment apparatus 200 are usually an IC (integrated circuit), an ASIC (application-limited integrated circuit), an LSI (large scale integrated circuit), a DSP (DSP). Digital signal processor) or a machine including a CPU-based processor and a PC.
- IC integrated circuit
- ASIC application-limited integrated circuit
- LSI large scale integrated circuit
- DSP digital signal processor
- Each of these modules can be provided in a plurality of single function LSIs or in one integrated LSI.
- the name used here is LSI, but it may also be called IC, system LSI, super LSI, or ultra LSI depending on the degree of integration.
- a method for realizing the integration not only the LSI but also a dedicated circuit or a general-purpose processor can be integrated.
- An FPGA Field Programmable Gate Array
- DSPs Digital Signal Processors
- An FPGA Field Programmable Gate Array
- LSI manufacture or a processor that can reconfigure the connection or configuration of the LSI can be used for the same application.
- manufacturing and processing technologies will improve and entirely new technologies may replace LSIs. Accumulation can also be performed by this technique.
- Stereoscopic display on a display device capable of displaying a stereoscopic image such as a liquid crystal display device or plasma display device that displays images in time series, a display device to which a lenticular lens layer is added, or another type of display device.
- the image alignment apparatus 200 may be incorporated.
- the stereoscopic image alignment device 200 may be implemented in digital media player devices such as digital video disc players, Blu-ray® disc players, and other types of digital media players. It is also possible to mount the stereoscopic image registration device 200 in other types of devices, and does not limit the scope of the present application.
- FIG. 3 is a block diagram showing a detailed functional configuration of the image alignment module 208 according to the present embodiment.
- the image registration module 208 includes a corresponding point detection unit 252, a first matrix calculation unit 254, a second matrix calculation unit 256, a matrix synthesis unit 258, a conversion unit 260, and a blurring unit.
- An area detection unit 220 and a viewing area calculation unit 262 are provided.
- the corresponding point detection unit 252 is a first point included in the first image that is one of the image pairs acquired by the image pair acquisition unit 205 and a point included in the second image that is the other. A corresponding point that is a set with a second point that is a point corresponding to this point is detected.
- the blur area detector 220 detects a blur area in the image pair. Any method for detecting the blur region can be used. For example, by dividing the image into a plurality of blocks and detecting the ratio of the portion where the density change of multiple pixels aligned in one direction is greater than or equal to the threshold for each block, blur region detection that identifies blocks with blur A method or the like may be used.
- the corresponding point detection unit 252 preferably detects the corresponding point from a region other than the region specified as the blur region in the image pair.
- the first matrix calculation unit 254 calculates a projective transformation matrix for converting the first point so that the vertical parallax between the first point and the second point is minimized and the epipolar constraint is satisfied. Note that the vertical parallax is a difference in vertical position values. Details will be described later.
- the second matrix calculation unit 256 has a third image coordinate system, which is the first image converted by the projective transformation matrix calculated by the first matrix calculation unit 254, has orthogonality, and the first matrix A slope transformation matrix for transforming the third image again is calculated so as to have the same aspect ratio as the image. More details will be described later.
- the viewing area calculation unit 262 calculates a viewing area that is an area common to the first image and the second image.
- the second matrix calculation unit 256 does not include all the third images, but the portion of the third image included in the viewing zone has orthogonality and the first image It is preferable to calculate the oblique transformation matrix so as to have the same aspect ratio as that of the portion included in the viewing area. Details of the viewing zone will be described later.
- the matrix composition unit 258 generates a composite transformation matrix by combining the projective transformation matrix and the oblique transformation matrix.
- the conversion unit 260 converts the first image using the composite conversion matrix.
- the image pair including the third image and the second image thus converted by the image alignment module 208 is output from the output unit 210 as a corrected image pair.
- FIG. 4A is a flowchart showing the operation of the image alignment module 208 according to the present embodiment.
- the image pair acquisition unit 205 acquires the image pair, so that the image pair is input to the stereoscopic image alignment apparatus 200.
- the corresponding point detection unit 252 detects the corresponding point of the image pair.
- step S304 is performed in order to shorten the processing time.
- the corresponding point detection unit 252 calculates the rotation parameter and the shift parameter used by the first matrix calculation unit 254 to create the projective transformation matrix so as to match the output from the gyro sensor that outputs information corresponding to the rotation angle. By doing so, the processing time for matrix calculation can be shortened and the accuracy can be improved.
- step S306 the first matrix calculation unit 254 calculates a basic matrix (F matrix) using the corresponding points.
- step S308 the first matrix calculation unit 254 calculates a projective transformation matrix using the corresponding points. This is intended to find a matrix for transforming one of the images so that the vertical parallax of the corresponding points is minimized and the epipolar constraint is satisfied.
- the second matrix calculation unit 256 calculates the oblique transformation matrix so that the viewing area of the deformed image maintains orthogonality and has the same aspect ratio as the viewing area before the deformation.
- the viewing area is a cut-out area used for stereoscopic viewing and includes an area common to both images. Only the corresponding points in the common area, which is the overlapping area of the left and right images, are useful for calculating the oblique transformation matrix, and the quality of the finally obtained 3D image is not guaranteed if other areas are used.
- step S312 the matrix synthesis unit 258 synthesizes the projective transformation matrix and the oblique transformation matrix.
- step S314 one image of the image pair is converted according to the combined conversion matrix (combined conversion matrix).
- step S316 the aligned stereoscopic image pair is output from the output unit 210.
- FIG. 4b is another flowchart including a blur region detection step S317 (b), a blur region compensation step S318 (b), a parallax adjustment step S319 (b), and a user feedback step S320 (b).
- step S304 (b) to detect corresponding points.
- a certain area is a target for corresponding point detection.
- the central region is preferable to use the central region as the target region for detecting the corresponding points in order to suitably represent the common or overlapping region of the two images.
- the blur area may be avoided from the target area of the corresponding point detection by using the result from the blur area detection step S317 (b). In order to reduce complexity or speed up feature point extraction, only an evenly distributed region around the center may be selected as a corresponding point detection target region.
- the corresponding point detection unit 252 selects a plurality of small regions uniformly distributed within a predetermined range from the center of each of the first image and the second image included in the image pair, Corresponding points may be detected from the small area. 12a and 12b show one such configuration, which will be described later in the present embodiment.
- step S306 (b), step S308 (b), and step S310 (b) by using only these selected candidate points, and excluding all points in the blur region, the first matrix calculation unit 254 and The second matrix calculation unit 256 generates a conversion matrix.
- step S318 (b) the blur region or blur object in one frame is compensated or shifted back by referring to the same blur region or step object in another frame. Then, the obtained image is subjected to deformation processing in step S314 (b).
- step S319 (b) the output unit 210 performs a parallax or 3D effect such as a pop-up effect (effect in which the image pops out) or a deep-in effect (effect in which the image sinks) from step S320 (b). Can be adjusted by changing the baseline or other parameters.
- a parallax or 3D effect such as a pop-up effect (effect in which the image pops out) or a deep-in effect (effect in which the image sinks) from step S320 (b).
- a parallax or 3D effect such as a pop-up effect (effect in which the image pops out) or a deep-in effect (effect in which the image sinks) from step S320 (b).
- a parallax or 3D effect such as a pop-up effect (effect in which the image pops out) or a deep-in effect (effect in which the image sinks) from step S320 (b).
- the output unit 210 may perform image processing on the image pair so that the 3
- user feedback can be performed by pressing a 3D display button, and a feedback signal is transmitted to the 3D processing engine shown in FIG.
- a feedback signal is transmitted to the 3D processing engine shown in FIG.
- pressing the “3D parallax” up / down button will make the depth deeper or shallower, or the button “pop-up” or “deep-in” Press to adjust the 3D effect of the content.
- the output unit 210 may acquire the size of the display device, and may output the image pair after performing image processing on the image pair so as to weaken the 3D effect as the display device is larger.
- the output unit 210 acquires adjustment information indicating at least one of the input operation from the viewing user, the size of the display device, and the distance between the display device and the viewing user, and based on the adjustment information, And adjusting and outputting at least one of the parallax, the 3D pop-up, and the deep-in effect of the third image and the second image.
- the output unit 210 adjusts the parallax between the left and right images included in the image pair. For example, in the cross parallax in which the right eye image is arranged on the left side of the left eye image with reference to the left eye image, the output unit 210 increases the cross parallax (that is, the right eye image). 3) The pop-up effect can be enhanced. Further, in the non-crossing parallax in which the right-eye image is arranged on the right side of the left-eye image with reference to the left-eye image, the larger the non-crossing parallax (that is, the right-eye image becomes the left-eye image). The deeper the effect is, the more the image is placed on the right side of the image.
- the corresponding point detection unit 252 may divide the image 1100 into five sub-blocks B1, B2, B3, B4, and B5 as shown in FIG. 12a. These blocks (or some blocks) may be used for feature detection, descriptor calculation, and corresponding point determination. In this case, it is not necessary to process the entire image. This speeds up the computation time of the system. In addition, these blocks help improve the accuracy of the transformation matrix and speed up the selection of the desired corresponding points. Also, in the case of a homogeneous block, there are usually no useful features. Therefore, homogeneous blocks should be excluded. Considering block B1 in FIG.
- B1 is a homogeneous block
- a non-homogeneous block around B1 eg, B1 ′ if B1 ′ is a non-homogeneous block
- B1 ′ if B1 ′ is a non-homogeneous block
- the block is homogeneous means that the dispersion of the image information included in the block is small. For example, (1) when the number of pixels (so-called edge portions) in which the pixel values of adjacent pixels in the block change more than a predetermined threshold value is smaller than a predetermined threshold value, or (2) block If the median or average value of each spatial frequency is smaller than a predetermined threshold value, the block may be determined to be homogeneous, and the other blocks may be determined to be non-homogeneous.
- Corresponding point detection unit 252 can use any corresponding point detection method in the prior art for detecting corresponding points. For example, each image in the image pair is divided into predetermined blocks having the same size, and the edges included in each block are extracted. After that, it is conceivable to calculate a similarity (for example, least square sum, likelihood, AIC, etc.) for each edge and detect a point included in the most similar edge as a corresponding point. More specifically, examples of known methods include “Object recognition from local scale-invariant features” in D.Lowe's "International Conference on Computer Vision” (1999), and H. Bay et al. It is disclosed in “SURF: Speeded up robustness” (Non-patent Document 2) in “European Conference on Computer Vision” (2006). Note that the corresponding point detection unit 252 may use other methods for detecting corresponding points, and this does not affect the scope and spirit of the present invention.
- a similarity for example, least square sum, likelihood, AIC, etc.
- step S306 the first matrix calculation unit 254 calculates a basic matrix F using the detected corresponding points.
- the method for obtaining the basic matrix F will be briefly described below.
- the 9-dimensional vector u may be estimated.
- the basic matrix F can be obtained by a method such as determining u numerically so as to maximize the likelihood by the maximum likelihood estimation method. More specifically, for example, R. Hartley and A. Zisserman, Multiple View Geometry in Computer Vision, Cambridge University Press, London, 2000 (Non-Patent Document 3), X. Armangue and J. Salvi, “Overall view regarding fundamental matrix estimation ", Image and Vision Computing, Vol. 21, 2003, pp. 205-220, and Z. Zhang,” Determining the epipolar geometry and its uncertainty: a review "," International Journal of Computer Vision " Vol. 27, 1998, pages 161-198 (Non-Patent Document 5) can be used.
- the second matrix calculation unit 256 calculates a projective transformation matrix for deforming one image of the image pair. Either image may be selected. In the present embodiment, the image for the right eye is selected as a deformation target. However, the selection of the left eye image does not affect the scope or spirit of the present invention. There may be an adaptive algorithm that selects a left-eye image or a right-eye image for deformation based on certain criteria. For example, the left eye image or the right eye image is more horizontal, the left eye image or the right eye image is clearer, or the left eye image or the right eye image. Such as which of the images contains the object of interest.
- the present invention uses projection transformation corresponding to epipolar geometry, that is, projection transformation that matches or matches the basic matrix (see Non-Patent Document 3). This is expressed by the following formula 2 in which the degree of freedom of normal projective transformation is limited to 3 instead of 8.
- F is the 3 ⁇ 3 basic matrix F calculated in step S306.
- [A] x b is an outer product, which means a ⁇ b.
- v T includes parameters of three degrees of freedom, a 1 ⁇ 3 vector.
- the first matrix calculation unit 254 as the vertical parallax between the corresponding points is minimized, obtaining a parameterized H by v T. That is, v satisfying the following formula 3 is obtained.
- XL is a list of corresponding points in the left image
- XR is a list of corresponding points in the right image.
- y_component_of (X) means the vertical coordinate component of point X.
- the first matrix calculation unit 254 can determine the parameter v T numerically from a plurality of corresponding points (X L and X R ). For example, it is possible to determine v that minimizes Equation 3 by the method of least squares.
- a method for obtaining H there is a method for obtaining H of each pair of corresponding points and obtaining an average thereof.
- the projective transformation matrix calculated in step S308 does not constrain parallax in the x direction. Therefore, since the image that has undergone projective transformation by H may be tilted, it is necessary to correct the tilt of the image by performing oblique tilt conversion. In order to calculate the tilt amount, the present invention restores the orthogonality of the aspect ratio and the viewing zone.
- view zone The terms “view zone”, “aspect ratio”, and “orthogonality” will be described with reference to FIGS. 6a and 6b.
- FIG. 6 a shows a left eye image 502 and a right eye image 504.
- Inner rectangles 506 and 508 indicate the outer shape of an area common to both images. The area outside the inner rectangles 506 and 508 appears only in one of the images. In stereoscopic view, these areas are uncomfortable. This is because these regions are viewed with only one eye, and the human brain cannot extract disparity information for perceiving depth.
- the rectangles 506 and 508 which are the inner common areas are called viewing areas.
- FIG. 6b shows a pre-deformation image 510 and a post-deformation image 512 of either the left or right image.
- the pre-deformation viewing area 514 has a certain aspect ratio, and this aspect ratio can be determined by the ratio between the distance between the points 518 and 522 and the distance between the points 520 and 524. Corner 526 is also a right angle.
- the post-deformation viewing zone 516 is determined by the deformation points 528, 530, 532, and 534 and may have different aspect ratios.
- the corner 540 may not be a right angle.
- a shear transformation matrix is calculated to restore orthogonality and aspect ratio.
- Equation 4 shows the general form of the shear transformation matrix.
- the matrix S expressed by Equation 4 represents the inclination of the right-eye image 352 that is the result of converting the right-eye image 350 by the matrix H calculated by the first matrix calculation unit 254. This is a matrix for returning to the original and obtaining the image 354 for the right eye whose coordinate axes are orthogonal.
- the corresponding points are as follows. That is, the point a of the right eye image 350 corresponds to the point a ′ of the right eye image 352, and the point a ′ of the right eye image 352 corresponds to the point a ′′ of the right eye image 354.
- a ′ Ha
- b ′ Hb
- c ′ Hc
- d ′ Hd
- a ′′ SHa
- b ′′ SHb
- c ′′ SHc
- d ′′ SHd.
- the line segment a ′′ c ′′ 355 and the line segment b ′′ d ′′ 356 in the right-eye image 354 are orthogonal to each other. Further, since the aspect ratios of the right eye image 350 and the right eye image 352 match, the ratio of the length of the line segment a ′′ c ′′ 355 to the length of the line segment b ′′ d ′′ 356 is The length h 0 of the line segment ac is equal to the ratio of the length w 0 of the line segment bd.
- the second matrix calculation unit 256 calculates the variables s1 and s2 included in the oblique transformation matrix S so as to satisfy both of the two constraints represented by the following Expression 5.
- the centroids of a plurality of corresponding points in the left eye image are calculated.
- corresponding points in the right-eye image are converted, and the centroids of the plurality of converted corresponding points are calculated.
- the difference between these centroids is calculated.
- the vertical component of the centroid difference corresponds to the vertical difference 536 in FIG. 6a.
- the horizontal component of the difference in the center of gravity corresponds to the horizontal difference 538 in FIG. 6a. Therefore, the size of the viewing zone can be derived from the difference between the original image size and the center of gravity.
- a perfect viewing zone may not be required. That is, there may be a difference between the calculated viewing area and the actual viewing area. However, satisfactory results are obtained with this method.
- Projective transformation matrix and shear transformation matrix are calculated using corresponding points.
- the image itself is not deformed. It is preferable to combine the projective transformation and the shear transformation matrix before performing one transformation process rather than two separate transformation processes.
- FIG. 14 is a conceptual diagram showing an image of image conversion processing using the projective transformation matrix H and the oblique transformation matrix S.
- the viewing area 362 included in the left-eye image plane and the viewing area 364a included in the right-eye image plane are images acquired by the image pair acquisition unit 205 according to the present embodiment.
- the viewing zones of the left and right images included in the pair are used.
- a point 372 included in the viewing area 362 and a point 378 included in the viewing area 364a are an image plane for the left eye and an image for the right eye when the point 370 is viewed from the left focus c ′ and the right focus c, respectively. It is a point that is copied on each surface. That is, point 372 and point 378 are corresponding points.
- the vertical positions of the point 372 and the point 378 are greatly different. Similarly, the vertical positions of other corresponding points are also greatly different.
- the viewing zone 364b is an image obtained by converting the viewing zone 364a by the projective transformation matrix H. By this conversion, the vertical positions of the corresponding points in the viewing area 362 and the viewing area 364b coincide. For example, the vertical position of the point 372 and the corresponding point 376 coincide with each other.
- the viewing zone 364b is inclined with the coordinate axes not orthogonal to each other.
- the viewing zone 364c is an image in which the vertical positions of the corresponding points (points 372 and 374) coincide with the viewing zone 362 and the orthogonality of the coordinates in the viewing zone is maintained.
- the viewing zone 364a can be directly converted to the viewing zone 364c.
- the first matrix calculation unit 254 calculates a projective transformation matrix H.
- the second matrix calculation unit 256 calculates the oblique transformation matrix S.
- the matrix synthesis unit 258 calculates the matrix H S.
- the conversion unit 260 performs conversion by the matrix H S on the viewing zone (for example, the viewing zone 364a) included in one of the acquired image pairs.
- the image alignment module 208 is incorporated in a photographing apparatus.
- the image alignment module 1246 in FIG. 7 has the same configuration as the stereoscopic image alignment module 200.
- FIG. 7 shows a schematic block diagram of a photographing apparatus 1200 according to another embodiment of the present invention.
- the photographing apparatus 1200 includes an optical system 1202, an image sensor 1204, an ADC (analog-digital converter) 1206, an image processor 1208, a microcomputer 1210, an external memory 1212, a driver controller 1220, and an OIS (optical type). (Shake prevention mechanism) sensor 1218, operation unit 1222, storage / transmission device 1216, and display device 1214.
- the image processor 1208 includes an internal memory 1240, an image registration module 1246, an original image processor 1242, and a color image processor 1244. Other components such as a microphone, speaker, etc. are not shown, but this does not limit the scope and spirit of the present invention.
- the optical system 1202 may include components such as a plurality of lenses or a set of lenses, a zoom / focus mechanism, an actuator, a shutter, and an opening in order to control an optical signal reaching the image sensor 1204.
- the image sensor 1204 stores an incident optical signal under the control of the microcomputer 1210 and converts the optical signal into an electrical signal.
- the electrical signal is converted into digital data (original image data) by the ADC 1206 and stored in the internal memory 1240 or the external memory 1212.
- the original image data may include a pair of stereoscopic images in which left and right images are continuously captured from different viewpoints.
- the original image processor 1242 acquires the original image data from the internal memory 1240 (or the external memory 1212), performs many preprocessing (not shown) such as noise reduction, linear correction, white balance, gamma correction, and the like.
- a preprocessed original image may be output. This preprocessed original image can be stored or transmitted by the storage / transmission device 1216.
- a color image such as RGB or YCbCr can be generated by processing the preprocessed original image by the color image processor 1244.
- the color image processor 1244 may include color interpolation, color correction, tonal range adjustment, color noise reduction, etc. to generate a suitable color image.
- the image registration module 1246 captures and aligns color images of a stereoscopic image pair.
- the output three-dimensional image pair that has been aligned can be displayed by a display device 1214 such as a liquid crystal monitor to which a lenticular lens layer capable of displaying a stereoscopic image for 3D viewing is attached.
- the image can also be displayed in 2D.
- the output registered stereo image pair can also be stored in the storage / transmission device.
- the preprocessed original image and output image may be compressed by a compression unit (not shown) before being stored / transmitted by the storage / transmission device 1216.
- Examples of storage devices include, but are not limited to, flash-based memory cards, hard drives, and optical drives.
- Examples of the transmission device include, but are not limited to, an HDMI interface, a USB interface, a wireless interface, and a direct-to-printer interface. Data processed by the storage device or the transmission device may be arbitrarily reversible or lossy compressed.
- the optical system 1202 may be controlled by the driver controller 1220.
- the driver controller 1220 is controlled by the microcomputer 1210.
- the operation unit 1222 receives a user operation input and transmits an electric signal to the microprocessor 1210, so that modules associated with the user input, such as a driver controller 1220, an image sensor 1204, an image processor 1208, and the like. Can be controlled.
- the OIS sensor 1218 detects shaking due to camera shake or camera movement, and transmits the electrical signal to the microcomputer 1210.
- the microcomputer 1210 controls the driver controller 1220 to control an actuator or the like in the optical system 1202 to move the lens so as to compensate for the shake, so that camera shake or camera shake is reduced.
- the details of the image alignment module 1246 can be referred to the embodiment described above. Processing by the image pair selection module 1251 may be performed prior to processing of the image alignment module 1246 to facilitate selection of image pairs that provide a better 3D effect that results in comfortable viewing on certain criteria. .
- the image processing 1208, the image alignment module 1246, and the modules inside thereof are usually realized in the form of an IC (Integrated Circuit), an ASIC (Application Specific Integrated Circuit), or an LSI (Large Scale Integrated Circuit).
- Each of these modules can be provided in a plurality of single function LSIs or in one integrated LSI.
- the name used here is LSI, but it may also be called IC, system LSI, super LSI, or ultra LSI depending on the degree of integration.
- LSI Integrated Circuit
- LSI Large Scale Integrated Circuit
- An FPGA Field Programmable Gate Array
- LSI manufacture or a processor that can reconfigure the connection or configuration of the LSI can be used for the same application.
- manufacturing and processing technologies will improve and entirely new technologies may replace LSIs. Accumulation can also be performed by this technique.
- FIG. 8 is an example of an image result illustrating the effect of the stereoscopic image alignment apparatus 200 according to Embodiment 1 of the present invention.
- the obtained image 13LR shows an overlay image between the left-eye image and the right-eye image of the stereoscopic image pair before image alignment. Further, a difference line (for example, L1302) indicating correspondence (displacement) between these images is shown superimposed on the overlay image. From the obtained image 13LR, it can be seen that the difference line extends long in both the vertical direction and the horizontal direction due to camera shake, and is not aligned.
- the stereoscopic image pair of the obtained image 13LR is displayed for 3D viewing, the viewer may experience eye strain and visual discomfort due to the displayed image.
- the obtained image 14LR shows an overlay image of the left image and the right image of the stereoscopic image pair after the image alignment according to the present invention is performed. From the obtained image 14LR, it can be seen that the difference line (L1304) is properly aligned (the image pair is preferably aligned). A line L1306 indicates a cutout area. When this aligned stereoscopic image pair is displayed, a very comfortable 3D view can be generated. Therefore, the effect of the present invention is demonstrated by FIG.
- the stereoscopic image registration device 200 may be used in an image capturing device that operates in the multi-shot mode. In such a mode, a plurality of images are automatically taken when the user presses a button. When the user moves the camera during multi-shot shooting, a plurality of images of the same scene are shot from different viewpoints.
- the image pair can be selected from a plurality of images. Selection criteria can be based on corresponding points. Preferably, image pairs with a sufficient number of corresponding points are selected for input to the image registration module.
- a method for detecting corresponding points is disclosed in the first embodiment.
- the time interval between each photographing can be set by a timer.
- Each photographing can be started by moving the photographing device. In order to determine such movement, a gyro sensor or an acceleration sensor can be used. It is preferable to start photographing when the photographing apparatus moves a predetermined distance so that the parallax is sufficient for stereoscopic viewing.
- the object may have different depths in one shooting scene.
- the distance for moving the apparatus When the object is close, the distance for moving the apparatus is short, and when the object is far, the distance for moving the apparatus is long.
- This distance can be calculated adaptively according to the depth of the object.
- One method for obtaining the depth is to obtain the depth based on the in-focus position.
- the stereoscopic image alignment apparatus 200 in a video shooting apparatus that creates an image pair by pairing the current image with one of adjacent images, the stereoscopic image alignment apparatus 200 according to Embodiment 1 can be used. Accordingly, a stereoscopic video can be generated from a normal 2D video.
- FIG. 9 shows possible uses.
- a normal camera or mobile phone is used to capture two or more images and is stored in an SD card or other memory card.
- the alignment of the 3D image can be performed inside the camera or the mobile phone.
- 3D image alignment can be performed in any player or display by incorporating the 3D image alignment module 208 into the player or display hardware.
- the alignment of the 3D image can also be performed by a PC or a server before being displayed on the 3D display.
- the feedback signal S29 in FIG. 2 may be the viewing distance from the viewer to the display.
- the feedback signal can be obtained automatically or via user interaction.
- the image alignment module 208 can adjust the baseline of the left and right images during or after image deformation or conversion.
- the API application interface
- an external button on the display screen is made like a button for adjusting the contrast or brightness of the display. Via this external button, the user can adjust visual comfort and 3D effects, and the user's request can be sent to the image registration module 208.
- the above configuration can be applied to other 3D display devices to create a user interface between the 3D processing module and the user.
- FIGS. 10a, 10b and 10c A detailed description of the different 3D effects is shown in FIGS. 10a, 10b and 10c.
- FIG. 10a the normal parallax is shown.
- a left object 9L that is an image for the left eye is displayed on the left side of the display screen 900
- a right object 9R that is an image for the right eye is displayed on the right side of the display screen 900.
- the observer perceives that the object PO is behind the display screen.
- x R is a position on the screen of the right-eye image
- x L is the position on the screen of the left-eye image.
- the perceived depth of the object can be controlled by d x .
- d x As d x increases, the perceived depth increases, while as d x decreases, the perceived depth decreases.
- FIG. 10b zero parallax is shown.
- the left object 9L and the right object 9R overlap each other and are displayed on the display screen 900. In this case, the observer perceives that the object PO is on the display screen.
- FIG. 10c negative parallax is shown.
- the left object 9L is displayed on the right side, and the right object 9R is displayed on the left side. In this case, the observer perceives that the object PO is in front of the display screen.
- d x is limited by the observer's eyes.
- the interocular space e is in the range of 50 mm to 70 mm (average 65 mm). This means that on the display screen, the physical difference between the left image and the right image must not exceed the distance e.
- the stereoscopic image alignment apparatus 200 has the same effect as the invention even if it does not include all of the above-described configurations.
- the stereoscopic image registration device 200 may not include the blur region detection unit 220. This is because the corresponding points can be detected from an image including blurring. In correspondence point detection, an image with less blur is preferable, so that, for example, in a shooting situation where camera shake is severe, improvement in the accuracy of correspondence point detection can be expected when the blur region detection unit 220 is provided.
- the stereoscopic image registration device 200 may not include the second matrix calculation unit 256.
- the conversion unit 260 stores a predetermined inclination conversion matrix and applies the inclination conversion matrix to the third image, so that the second matrix calculation unit 256 is not provided.
- the second matrix calculation unit 256 calculates a slope transformation matrix according to the shooting conditions, thereby making it possible to impart a more natural stereoscopic effect to the image.
- the conversion unit 260 converts the first image using the projective transformation matrix calculated by the first matrix calculation unit 254. .
- the stereoscopic image registration device 200 may not include the viewing area calculation unit 262. It is generally considered that areas common to the left and right images do not differ greatly each time. Therefore, for example, the second matrix is obtained by following a predetermined rule, such as specifying the region for 80% of the left eye image and the region of the left 80% portion of the right eye image as the viewing region.
- the calculation unit 256 can determine the processing region in the image without the viewing zone calculation unit 262. By providing the viewing area calculation unit 262, the viewing area can be specified more accurately, and as a result, a more natural stereoscopic effect can be added to the image.
- the stereoscopic image registration device 200 may not include the matrix composition unit 258.
- the matrix synthesis unit 258 is not necessary.
- the conversion unit 260 sequentially uses the projective transformation matrix and the oblique transformation matrix stored in the storage area. Since conversion processing can be performed on one image, the same invention effect can be obtained without performing synthesis by the matrix synthesis unit 258.
- the image pair acquisition unit 205 acquires (A) an image pair candidate that is a candidate for a plurality of image pairs, and (B) another image pair candidate for each of the plurality of image pair candidates.
- a common visual field that is a common visual field with each of the information, obtain quality information that is information on at least one of the light amount, luminance, color, and sharpness in the common visual field, and (C) according to the quality information
- An image pair candidate with the best image quality shown may be selected as an image pair.
- (A) above as a method of acquiring a plurality of image pairs, for example, it is easily acquired by holding a digital still camera set to the continuous shooting mode and moving the user's holding hand in the horizontal direction. Is possible.
- a processor such as a CPU executing a program.
- the present invention may be the above program or a recording medium on which the above program is recorded.
- the program can be distributed via a transmission medium such as the Internet.
- the present invention can be applied to a stereoscopic image alignment device that detects corresponding points from a plurality of image pairs and aligns them so as to satisfy a certain standard, and in particular, stereoscopic viewing is possible in a stereoscopic display device.
- the present invention can be applied to a stereoscopic image alignment device that aligns image pairs in such a manner.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
- Image Processing (AREA)
- Studio Devices (AREA)
Abstract
Description
102、402、406 左画像
104、404、408、410 右画像
106 単眼式デジタルカメラ
108 距離
200 立体視用画像位置合わせ装置
202 記憶媒体リーダ
204 画像デコーダ
205 画像対取得部
208 画像位置合わせモジュール
210 出力部
212 内部バッファ
214 表示装置
216 記憶/送信装置
218 プリンタ
220 ブレ領域検出部
252 対応点検出部
254 第1行列算出部
256 第2行列算出部
258 行列合成部
260 変換部
262 視域算出部
350、352、354、504 右眼用画像
355 線分a’’c’’
356 線分b’’d’’
362、364a、364b、364c 視域
370、372、374、376、378 点
502 左眼用画像
506、508 矩形
510 変形前画像
512 変形後画像
514 変形前視域
516 変形後視域
518、520、522、524 点
528、530、532、534 変形点
526、540 角
536 垂直差
538 水平差
900 表示画面
L1302、L1304 差分線
Claims (17)
- 左眼用画像と、前記左眼用画像に対応する右眼用画像とを含む画像対を取得する画像対取得部と、
前記画像対の一方である第1の画像に含まれる第1の点と、他方である第2の画像に含まれる点であり、前記第1の点に対応する点である第2の点との組である対応点を検出する対応点検出部と、
前記第1の点と前記第2の点との垂直視差が最小となり、かつエピポーラ拘束を満たすように前記第1の点を変換する射影変換行列を算出する第1行列算出部と、
前記射影変換行列を用いて前記第1の画像を変換する変換部と、
前記変換された第1の画像である第3の画像と、前記第2の画像とを出力する出力部とを備える
立体視用画像位置合わせ装置。 - 前記画像対取得部は、2つのイメージセンサと2つのレンズ、1つのイメージセンサと2つのレンズ、または2つのイメージセンサと1つのレンズにより同時に撮影された2つの画像を前記画像対として取得する
請求項1に記載の立体視用画像位置合わせ装置。 - 前記画像対取得部は、1つのイメージセンサと1つのレンズを用いて、異なる視点から異なる時間に撮影された2つの画像を前記画像対として取得する
請求項1に記載の立体視用画像位置合わせ装置。 - さらに、前記第3の画像の座標系が直交性を有し、かつ、前記第1の画像と同じアスペクト比を有するように前記第3の画像を変換する斜傾変換行列を算出する第2行列算出部と、
前記射影変換行列と前記斜傾変換行列とを組み合わせて合成変換行列を生成する行列合成部とを備え、
前記変換部は、前記合成変換行列を用いて、前記第1の画像を変換する
請求項1に記載の立体視用画像位置合わせ装置。 - さらに、前記第1の画像及び前記第2の画像に共通する領域である視域を算出する視域算出部を備え、
前記第2行列算出部は、前記第3の画像のうち前記視域に含まれる部分の座標系が、直交性を有し、かつ、前記第1の画像のうち前記視域に含まれる部分と同じアスペクト比を有するように前記斜傾変換行列を算出する
請求項4に記載の立体視用画像位置合わせ装置。 - 前記対応点検出部は、前記画像対に含まれる前記第1の画像及び前記第2の画像の各々の中心から事前に定められた範囲内において、均一に分散された複数の小領域を選択し、当該小領域のなかから前記対応点を検出する
請求項1に記載の立体視用画像位置合わせ装置。 - さらに、前記画像対においてブレ領域を検出するブレ領域検出部を備え、
前記対応点検出部は、前記画像対のうち前記ブレ領域以外の領域から前記対応点を検出する
請求項1または3に記載の立体視用画像位置合わせ装置。 - 前記出力部は、視聴ユーザからの入力操作、表示装置の大きさ、及び、表示装置と視聴ユーザとの距離のうち少なくとも1つを示す調整情報を取得し、前記調整情報に基づいて、前記第3の画像及び前記第2の画像の視差、3Dポップアップ、及びディープイン効果のうち少なくとも1つを調整して出力する
請求項1に記載の立体視用画像位置合わせ装置。 - 前記画像対取得部は、(A)複数の画像対の候補である画像対候補を取得し、(B)前記複数の画像対候補の各々について、他の画像対候補の各々と共通する前記視域である共通視域を特定し、前記共通視域における光量、輝度、色、及び鮮明度のうち少なくとも1つに関する情報である品質情報を取得し、(C)前記品質情報によって示される画像品質が最もよい前記画像対候補を前記画像対として選択する
請求項1に記載の立体視用画像位置合わせ装置。 - 前記画像対取得部は、(1)前記左眼用画像及び右眼用画像を含む画像群が圧縮され、かつメモリカードに記憶されている場合、または、(2)前記画像群が圧縮され、かつ外部装置からネットワークを介して送信されている場合は、圧縮されている前記画像群を復号する
請求項1に記載の立体視用画像位置合わせ装置。 - 前記対応点検出部は、回転角に対応する情報を出力するジャイロセンサからの出力に一致するように、前記射影変換行列の作成に用いる回転パラメータおよびシフトパラメータを算出する
請求項1に記載の立体視用画像位置合わせ装置。 - 2Dカメラに組み込むことで3Dカメラになり、3D処理エンジンとして3Dディスプレイに組み込むことで2D画像群を3D画像または3D映像に変換し、プレーヤまたはレコーダに組み込むことで2D画像群を3D画像または3D映像に変換する
請求項1に記載の立体視用画像位置合わせ装置。 - 左眼用画像と、前記左眼用画像に対応する右眼用画像とを含む画像対を取得し、
前記画像対の一方である第1の画像に含まれる第1の点と、他方である第2の画像に含まれる点であり、前記第1の点に対応する点である第2の点との組である対応点を検出し、
前記第1の点と、前記第2の点とが同じ垂直位置値を有し、かつエピポーラ拘束を満たすように前記第1の点を変換する射影変換行列を算出し、
前記射影変換行列を用いて前記第1の画像を変換し、
前記変換された第1の画像である第3の画像と、前記第2の画像とを出力する
立体視用画像位置合わせ方法。 - 請求項13に記載の立体視用画像位置合わせ方法をコンピュータに実行させる
プログラム。 - 請求項14に記載のプログラムを記録した
コンピュータ読み取り可能な記録媒体。 - 左眼用画像と右眼用画像とを含む画像対を取得する画像対取得部と、
前記画像対に含まれる一方である第1の画像に含まれる第1の点と、他方である第2の画像に含まれる点であり、前記第1の点に対応する点である第2の点との組である対応点を検出する対応点検出部と、
前記第1の点と、前記第2の点とが同じ垂直位置値を有し、かつエピポーラ拘束を満たすように前記第1の点を変換する射影変換行列を算出する第1行列算出部と、
前記射影変換行列を用いて前記第1の画像を変換する変換部と、
前記変換された第1の画像である第3の画像と、前記第2の画像とを出力する出力部とを備える
集積回路。 - 請求項3に記載の立体視用画像位置合わせ装置を備える
デジタルスチルカメラ。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/320,640 US8890934B2 (en) | 2010-03-19 | 2011-03-14 | Stereoscopic image aligning apparatus, stereoscopic image aligning method, and program of the same |
JP2011536681A JP5679978B2 (ja) | 2010-03-19 | 2011-03-14 | 立体視用画像位置合わせ装置、立体視用画像位置合わせ方法、及びそのプログラム |
CN201180002079.6A CN102428707B (zh) | 2010-03-19 | 2011-03-14 | 立体视用图像对位装置和立体视用图像对位方法 |
EP11755876.7A EP2549762B1 (en) | 2010-03-19 | 2011-03-14 | Stereovision-image position matching apparatus, stereovision-image position matching method, and program therefor |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2010-065185 | 2010-03-19 | ||
JP2010065185 | 2010-03-19 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2011114683A1 true WO2011114683A1 (ja) | 2011-09-22 |
Family
ID=44648797
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2011/001455 WO2011114683A1 (ja) | 2010-03-19 | 2011-03-14 | 立体視用画像位置合わせ装置、立体視用画像位置合わせ方法、及びそのプログラム |
Country Status (5)
Country | Link |
---|---|
US (1) | US8890934B2 (ja) |
EP (1) | EP2549762B1 (ja) |
JP (1) | JP5679978B2 (ja) |
CN (1) | CN102428707B (ja) |
WO (1) | WO2011114683A1 (ja) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130127843A1 (en) * | 2011-11-22 | 2013-05-23 | Samsung Electronics Co., Ltd. | Display apparatus and display method thereof |
EP2761875A1 (en) * | 2011-09-29 | 2014-08-06 | Qualcomm Incorporated | Methods and apparatus for conditional display of a stereoscopic image pair |
JP2015502058A (ja) * | 2011-11-04 | 2015-01-19 | クゥアルコム・インコーポレイテッドQualcomm Incorporated | マルチスペクトル撮像システム |
JP2017005380A (ja) * | 2015-06-05 | 2017-01-05 | キヤノン株式会社 | 制御装置、撮像装置、制御方法、プログラム、および、記憶媒体 |
US9961322B2 (en) | 2011-12-09 | 2018-05-01 | Lg Innotek Co., Ltd. | Apparatus and method for eliminating noise in stereo image |
CN110223355A (zh) * | 2019-05-15 | 2019-09-10 | 大连理工大学 | 一种基于双重极线约束的特征标志点匹配方法 |
CN110969659A (zh) * | 2019-10-31 | 2020-04-07 | 浙江未来技术研究院(嘉兴) | 一种被动式标记点的空间定位装置和方法 |
Families Citing this family (94)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11792538B2 (en) | 2008-05-20 | 2023-10-17 | Adeia Imaging Llc | Capturing and processing of images including occlusions focused on an image sensor by a lens stack array |
US8866920B2 (en) | 2008-05-20 | 2014-10-21 | Pelican Imaging Corporation | Capturing and processing of images using monolithic camera array with heterogeneous imagers |
EP2289235A4 (en) | 2008-05-20 | 2011-12-28 | Pelican Imaging Corp | RECORDING AND PROCESSING IMAGES BY MONOLITHIC CAMERA ARRANGEMENT WITH HETEROGENIC IMAGE TRANSFORMER |
WO2011063347A2 (en) | 2009-11-20 | 2011-05-26 | Pelican Imaging Corporation | Capturing and processing of images using monolithic camera array with heterogeneous imagers |
SG10201503516VA (en) | 2010-05-12 | 2015-06-29 | Pelican Imaging Corp | Architectures for imager arrays and array cameras |
US8878950B2 (en) | 2010-12-14 | 2014-11-04 | Pelican Imaging Corporation | Systems and methods for synthesizing high resolution images using super-resolution processes |
EP2509324A1 (en) * | 2011-04-08 | 2012-10-10 | Thomson Licensing | Method and apparatus for analyzing stereoscopic or multi-view images |
WO2012155119A1 (en) | 2011-05-11 | 2012-11-15 | Pelican Imaging Corporation | Systems and methods for transmitting and receiving array camera image data |
US20130265459A1 (en) | 2011-06-28 | 2013-10-10 | Pelican Imaging Corporation | Optical arrangements for use with an array camera |
EP2726930A4 (en) | 2011-06-28 | 2015-03-04 | Pelican Imaging Corp | OPTICAL ARRANGEMENTS FOR USE WITH AN ARRAY CAMERA |
US20130033713A1 (en) * | 2011-08-02 | 2013-02-07 | Samsung Electronics Co., Ltd | Apparatus and method of forming image, terminal and method of print control, and computer-readable medium |
US9305361B2 (en) * | 2011-09-12 | 2016-04-05 | Qualcomm Incorporated | Resolving homography decomposition ambiguity based on orientation sensors |
WO2013043761A1 (en) | 2011-09-19 | 2013-03-28 | Pelican Imaging Corporation | Determining depth from multiple views of a scene that include aliasing using hypothesized fusion |
KR102002165B1 (ko) | 2011-09-28 | 2019-07-25 | 포토내이션 리미티드 | 라이트 필드 이미지 파일의 인코딩 및 디코딩을 위한 시스템 및 방법 |
EP2817955B1 (en) | 2012-02-21 | 2018-04-11 | FotoNation Cayman Limited | Systems and methods for the manipulation of captured light field image data |
US9210392B2 (en) | 2012-05-01 | 2015-12-08 | Pelican Imaging Coporation | Camera modules patterned with pi filter groups |
KR20150023907A (ko) | 2012-06-28 | 2015-03-05 | 펠리칸 이매징 코포레이션 | 결함있는 카메라 어레이들, 광학 어레이들 및 센서들을 검출하기 위한 시스템들 및 방법들 |
US20140002674A1 (en) | 2012-06-30 | 2014-01-02 | Pelican Imaging Corporation | Systems and Methods for Manufacturing Camera Modules Using Active Alignment of Lens Stack Arrays and Sensors |
US20140195983A1 (en) * | 2012-06-30 | 2014-07-10 | Yangzhou Du | 3d graphical user interface |
CN107346061B (zh) | 2012-08-21 | 2020-04-24 | 快图有限公司 | 用于使用阵列照相机捕捉的图像中的视差检测和校正的系统和方法 |
WO2014032020A2 (en) | 2012-08-23 | 2014-02-27 | Pelican Imaging Corporation | Feature based high resolution motion estimation from low resolution images captured using an array source |
US8948497B2 (en) * | 2012-09-04 | 2015-02-03 | Digital Signal Corporation | System and method for increasing resolution of images obtained from a three-dimensional measurement system |
US9214013B2 (en) | 2012-09-14 | 2015-12-15 | Pelican Imaging Corporation | Systems and methods for correcting user identified artifacts in light field images |
EP2901671A4 (en) | 2012-09-28 | 2016-08-24 | Pelican Imaging Corp | CREATING IMAGES FROM LIGHT FIELDS USING VIRTUAL POINTS OF VIEW |
US9148651B2 (en) * | 2012-10-05 | 2015-09-29 | Blackberry Limited | Methods and devices for generating a stereoscopic image |
CN103792667B (zh) | 2012-10-30 | 2016-06-01 | 财团法人工业技术研究院 | 立体摄像装置、自动校正装置与校正方法 |
US9143711B2 (en) | 2012-11-13 | 2015-09-22 | Pelican Imaging Corporation | Systems and methods for array camera focal plane control |
EP2757524B1 (en) * | 2013-01-16 | 2018-12-19 | Honda Research Institute Europe GmbH | Depth sensing method and system for autonomous vehicles |
WO2014130849A1 (en) | 2013-02-21 | 2014-08-28 | Pelican Imaging Corporation | Generating compressed light field representation data |
US9374512B2 (en) | 2013-02-24 | 2016-06-21 | Pelican Imaging Corporation | Thin form factor computational array cameras and modular array cameras |
US9638883B1 (en) | 2013-03-04 | 2017-05-02 | Fotonation Cayman Limited | Passive alignment of array camera modules constructed from lens stack arrays and sensors based upon alignment information obtained during manufacture of array camera modules using an active alignment process |
WO2014138697A1 (en) | 2013-03-08 | 2014-09-12 | Pelican Imaging Corporation | Systems and methods for high dynamic range imaging using array cameras |
US8866912B2 (en) | 2013-03-10 | 2014-10-21 | Pelican Imaging Corporation | System and methods for calibration of an array camera using a single captured image |
WO2014165244A1 (en) | 2013-03-13 | 2014-10-09 | Pelican Imaging Corporation | Systems and methods for synthesizing images from image data captured by an array camera using restricted depth of field depth maps in which depth estimation precision varies |
US9124831B2 (en) | 2013-03-13 | 2015-09-01 | Pelican Imaging Corporation | System and methods for calibration of an array camera |
US9106784B2 (en) | 2013-03-13 | 2015-08-11 | Pelican Imaging Corporation | Systems and methods for controlling aliasing in images captured by an array camera for use in super-resolution processing |
WO2014164909A1 (en) | 2013-03-13 | 2014-10-09 | Pelican Imaging Corporation | Array camera architecture implementing quantum film sensors |
US9100586B2 (en) | 2013-03-14 | 2015-08-04 | Pelican Imaging Corporation | Systems and methods for photometric normalization in array cameras |
US9578259B2 (en) | 2013-03-14 | 2017-02-21 | Fotonation Cayman Limited | Systems and methods for reducing motion blur in images or video in ultra low light with array cameras |
US9633442B2 (en) | 2013-03-15 | 2017-04-25 | Fotonation Cayman Limited | Array cameras including an array camera module augmented with a separate camera |
US10122993B2 (en) | 2013-03-15 | 2018-11-06 | Fotonation Limited | Autofocus system for a conventional camera that uses depth information from an array camera |
US9497370B2 (en) | 2013-03-15 | 2016-11-15 | Pelican Imaging Corporation | Array camera architecture implementing quantum dot color filters |
US9445003B1 (en) | 2013-03-15 | 2016-09-13 | Pelican Imaging Corporation | Systems and methods for synthesizing high resolution images using image deconvolution based on motion and depth information |
US9497429B2 (en) | 2013-03-15 | 2016-11-15 | Pelican Imaging Corporation | Extended color processing on pelican array cameras |
WO2014145856A1 (en) | 2013-03-15 | 2014-09-18 | Pelican Imaging Corporation | Systems and methods for stereo imaging with camera arrays |
CN105339756B (zh) * | 2013-04-08 | 2019-08-06 | C3D增强实感解决方案有限公司 | 使用多个相机设备的距离评估 |
JP2014207519A (ja) * | 2013-04-11 | 2014-10-30 | ソニー株式会社 | 画像処理装置、画像処理方法、プログラム、および電子機器 |
CN104113684B (zh) * | 2013-04-15 | 2017-09-22 | 宏达国际电子股份有限公司 | 控制方法及电子装置 |
US9898856B2 (en) | 2013-09-27 | 2018-02-20 | Fotonation Cayman Limited | Systems and methods for depth-assisted perspective distortion correction |
CN103996215A (zh) * | 2013-11-05 | 2014-08-20 | 深圳市云立方信息科技有限公司 | 一种实现虚拟视图转立体视图的方法及装置 |
WO2015070105A1 (en) | 2013-11-07 | 2015-05-14 | Pelican Imaging Corporation | Methods of manufacturing array camera modules incorporating independently aligned lens stacks |
WO2015074078A1 (en) | 2013-11-18 | 2015-05-21 | Pelican Imaging Corporation | Estimating depth from projected texture using camera arrays |
EP3075140B1 (en) | 2013-11-26 | 2018-06-13 | FotoNation Cayman Limited | Array camera configurations incorporating multiple constituent array cameras |
CN104811688B (zh) * | 2014-01-28 | 2017-09-01 | 聚晶半导体股份有限公司 | 图像获取装置及其图像形变检测方法 |
TWI520098B (zh) | 2014-01-28 | 2016-02-01 | 聚晶半導體股份有限公司 | 影像擷取裝置及其影像形變偵測方法 |
US10089740B2 (en) | 2014-03-07 | 2018-10-02 | Fotonation Limited | System and methods for depth regularization and semiautomatic interactive matting using RGB-D images |
US9247117B2 (en) | 2014-04-07 | 2016-01-26 | Pelican Imaging Corporation | Systems and methods for correcting for warpage of a sensor array in an array camera module by introducing warpage into a focal plane of a lens stack array |
GB2525170A (en) | 2014-04-07 | 2015-10-21 | Nokia Technologies Oy | Stereo viewing |
US9521319B2 (en) | 2014-06-18 | 2016-12-13 | Pelican Imaging Corporation | Array cameras and array camera modules including spectral filters disposed outside of a constituent image sensor |
US9613408B2 (en) * | 2014-09-25 | 2017-04-04 | Intel Corporation | High dynamic range image composition using multiple images |
CN113256730B (zh) | 2014-09-29 | 2023-09-05 | 快图有限公司 | 用于阵列相机的动态校准的系统和方法 |
KR101932547B1 (ko) * | 2014-10-23 | 2018-12-27 | 한화테크윈 주식회사 | 카메라 시스템 및 이의 영상 정합 방법 |
HUP1400600A2 (hu) * | 2014-12-17 | 2016-06-28 | Pi Holding Zrt | Eljárás képszegmens tartalmának lecserélésére |
US10931933B2 (en) * | 2014-12-30 | 2021-02-23 | Eys3D Microelectronics, Co. | Calibration guidance system and operation method of a calibration guidance system |
EP3278709B1 (en) * | 2015-03-31 | 2020-01-08 | Sony Corporation | Medical observation device, information processing method, program and video microscope device |
US10275863B2 (en) * | 2015-04-03 | 2019-04-30 | Cognex Corporation | Homography rectification |
US9942474B2 (en) | 2015-04-17 | 2018-04-10 | Fotonation Cayman Limited | Systems and methods for performing high speed video capture and depth estimation using array cameras |
JP6628866B2 (ja) * | 2015-09-08 | 2020-01-15 | エスゼット ディージェイアイ テクノロジー カンパニー リミテッドSz Dji Technology Co.,Ltd | 三次元表示を実現する方法、及びシステム |
US10097747B2 (en) * | 2015-10-21 | 2018-10-09 | Qualcomm Incorporated | Multiple camera autofocus synchronization |
US9762893B2 (en) * | 2015-12-07 | 2017-09-12 | Google Inc. | Systems and methods for multiscopic noise reduction and high-dynamic range |
CN109070803B (zh) * | 2016-04-14 | 2021-10-08 | 金泰克斯公司 | 提供深度信息的车辆显示系统 |
US10796425B1 (en) * | 2016-09-06 | 2020-10-06 | Amazon Technologies, Inc. | Imagery-based member deformation gauge |
KR102529928B1 (ko) * | 2016-09-22 | 2023-05-09 | 삼성전자주식회사 | 스테레오 카메라의 교정 방법 및 이를 수행하는 전자 장치 |
US10284761B2 (en) | 2016-11-17 | 2019-05-07 | Motorola Mobility Llc | Multi-camera capture of a high dynamic range image |
CN107046638A (zh) * | 2016-12-30 | 2017-08-15 | 无锡易维视显示技术有限公司 | 单摄像头的3d影像拍摄方法 |
US10250794B2 (en) | 2017-01-04 | 2019-04-02 | Motorola Mobility Llc | Capturing an image using multi-camera automatic focus |
US10169671B2 (en) | 2017-02-17 | 2019-01-01 | Motorola Mobility Llc | Face detection with temperature and distance validation |
US10250795B2 (en) | 2017-03-15 | 2019-04-02 | Motorola Mobility Llc | Identifying a focus point in a scene utilizing a plurality of cameras |
CN107147898B (zh) * | 2017-05-10 | 2019-02-05 | 宁波视睿迪光电有限公司 | 一种3d显示屏对位系统和方法 |
US10482618B2 (en) | 2017-08-21 | 2019-11-19 | Fotonation Limited | Systems and methods for hybrid depth regularization |
WO2020097128A1 (en) | 2018-11-06 | 2020-05-14 | Flir Commercial Systems, Inc. | Automatic co-registration of thermal and visible image pairs |
US11650596B2 (en) * | 2019-05-31 | 2023-05-16 | Cascade Corporation | Load alignment aid |
CN110533711A (zh) * | 2019-09-04 | 2019-12-03 | 云南电网有限责任公司带电作业分公司 | 一种基于加速稳健特征的高效大尺度立体匹配算法 |
KR102646521B1 (ko) | 2019-09-17 | 2024-03-21 | 인트린식 이노베이션 엘엘씨 | 편광 큐를 이용한 표면 모델링 시스템 및 방법 |
EP4042101A4 (en) | 2019-10-07 | 2023-11-22 | Boston Polarimetrics, Inc. | SYSTEMS AND METHODS FOR DETECTING SURFACE NORMALS USING POLARIZATION |
EP4066001A4 (en) | 2019-11-30 | 2024-01-24 | Boston Polarimetrics, Inc. | SYSTEMS AND METHODS FOR TRANSPARENT OBJECT SEGMENTATION USING POLARIZATION GUIDES |
US11375180B2 (en) * | 2019-12-10 | 2022-06-28 | Yousseff El Moumni | Video engine for generating an enhanced impression of depth based upon two-dimensioned video content |
JP7462769B2 (ja) | 2020-01-29 | 2024-04-05 | イントリンジック イノベーション エルエルシー | 物体の姿勢の検出および測定システムを特徴付けるためのシステムおよび方法 |
WO2021154459A1 (en) | 2020-01-30 | 2021-08-05 | Boston Polarimetrics, Inc. | Systems and methods for synthesizing data for training statistical models on different imaging modalities including polarized images |
US11953700B2 (en) | 2020-05-27 | 2024-04-09 | Intrinsic Innovation Llc | Multi-aperture polarization optical systems using beam splitters |
US12020455B2 (en) | 2021-03-10 | 2024-06-25 | Intrinsic Innovation Llc | Systems and methods for high dynamic range image reconstruction |
US11954886B2 (en) | 2021-04-15 | 2024-04-09 | Intrinsic Innovation Llc | Systems and methods for six-degree of freedom pose estimation of deformable objects |
US11290658B1 (en) | 2021-04-15 | 2022-03-29 | Boston Polarimetrics, Inc. | Systems and methods for camera exposure control |
US11689813B2 (en) | 2021-07-01 | 2023-06-27 | Intrinsic Innovation Llc | Systems and methods for high dynamic range imaging using crossed polarizers |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6191809B1 (en) | 1998-01-15 | 2001-02-20 | Vista Medical Technologies, Inc. | Method and apparatus for aligning stereo images |
JP2002324234A (ja) * | 2001-02-23 | 2002-11-08 | Sharp Corp | 立体画像を偏歪修正する方法および装置 |
US6606404B1 (en) | 1999-06-19 | 2003-08-12 | Microsoft Corporation | System and method for computing rectifying homographies for stereo vision processing of three dimensional objects |
JP2003244500A (ja) | 2002-02-13 | 2003-08-29 | Pentax Corp | ステレオ画像撮像装置 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7343035B1 (en) * | 2003-10-20 | 2008-03-11 | Open Invention Network Llc | Method and system for three-dimensional feature attribution through synergy of rational polynomial coefficients and projective geometry |
DE102006050379A1 (de) * | 2006-10-25 | 2008-05-08 | Norbert Prof. Dr. Link | Verfahren und Vorrichtung zur Überwachung eines Raumvolumens sowie Kalibrierverfahren |
US8135238B2 (en) * | 2008-06-05 | 2012-03-13 | Kia Sha Managment Liability Company | Free view generation in ray-space |
CN102576154A (zh) * | 2009-10-30 | 2012-07-11 | 惠普发展公司,有限责任合伙企业 | 立体显示系统 |
-
2011
- 2011-03-14 CN CN201180002079.6A patent/CN102428707B/zh not_active Expired - Fee Related
- 2011-03-14 WO PCT/JP2011/001455 patent/WO2011114683A1/ja active Application Filing
- 2011-03-14 JP JP2011536681A patent/JP5679978B2/ja not_active Expired - Fee Related
- 2011-03-14 EP EP11755876.7A patent/EP2549762B1/en not_active Not-in-force
- 2011-03-14 US US13/320,640 patent/US8890934B2/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6191809B1 (en) | 1998-01-15 | 2001-02-20 | Vista Medical Technologies, Inc. | Method and apparatus for aligning stereo images |
JP2001517414A (ja) * | 1998-01-15 | 2001-10-02 | ビスタ・メディカル・テクノロジーズ・インコーポレーテッド | 立体像を位置合わせする方法および装置 |
US6606404B1 (en) | 1999-06-19 | 2003-08-12 | Microsoft Corporation | System and method for computing rectifying homographies for stereo vision processing of three dimensional objects |
JP2002324234A (ja) * | 2001-02-23 | 2002-11-08 | Sharp Corp | 立体画像を偏歪修正する方法および装置 |
US7113632B2 (en) | 2001-02-23 | 2006-09-26 | Sharp Kabushiki Kaisha | Method of and apparatus for rectifying a stereoscopic image |
JP2003244500A (ja) | 2002-02-13 | 2003-08-29 | Pentax Corp | ステレオ画像撮像装置 |
Non-Patent Citations (11)
Title |
---|
D. LOWE: "Object recognition from local scale-invariant features", INTERNATIONAL CONFERENCE ON COMPUTER VISION, 1999 |
D.LOWE: "Object recognition from local scale-invariant features", INTERNATIONAL CONFERENCE ON COMPUTER VISION, 1999 |
H. BAY ET AL.: "SURF: Speeded up robust features", EUROPEAN CONFERENCE ON COMPUTER VISION, 2006 |
H.BAY ET AL.: "SURF: Speeded up robust features", EUROPEAN CONFERENCE ON COMPUTER VISION, 2006 |
R. HARTLEY, A. ZISSERMAN: "Multiple View Geometry in Computer Vision", 2000, CAMBRIDGE UNIVERSITY PRESS |
R.HARTLEY, A.ZISSERMAN: "Multiple View Geometry in Computer Vision", 2000, CAMBRIDGE UNIVERSITY PRESS |
See also references of EP2549762A4 |
X. ARMANGUE, J. SALVI: "Overall view regarding fundamental matrix estimation", IMAGE AND VISION COMPUTING, vol. 21, 2003, pages 205 - 220, XP055284449, DOI: doi:10.1016/S0262-8856(02)00154-3 |
X.ARMANGUE, J. SALVI: "Overall view regarding fundamental matrix estimation", IMAGE AND VISION COMPUTING, vol. 21, 2003, pages 205 - 220, XP055284449, DOI: doi:10.1016/S0262-8856(02)00154-3 |
Z. ZHANG: "Determining the epipolar geometry and its uncertainty: a review", INTERNATIONAL JOURNAL OF COMPUTER VISION, vol. 27, 1998, pages 161 - 198 |
Z.ZHANG: "Determining the epipolar geometry and its uncertainty: a review", INTERNATIONAL JOURNAL OF COMPUTER VISION, vol. 27, 1998, pages 161 - 198 |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2761875A1 (en) * | 2011-09-29 | 2014-08-06 | Qualcomm Incorporated | Methods and apparatus for conditional display of a stereoscopic image pair |
JP2015502058A (ja) * | 2011-11-04 | 2015-01-19 | クゥアルコム・インコーポレイテッドQualcomm Incorporated | マルチスペクトル撮像システム |
US9692991B2 (en) | 2011-11-04 | 2017-06-27 | Qualcomm Incorporated | Multispectral imaging system |
US20130127843A1 (en) * | 2011-11-22 | 2013-05-23 | Samsung Electronics Co., Ltd. | Display apparatus and display method thereof |
US9547933B2 (en) * | 2011-11-22 | 2017-01-17 | Samsung Electronics Co., Ltd | Display apparatus and display method thereof |
US9961322B2 (en) | 2011-12-09 | 2018-05-01 | Lg Innotek Co., Ltd. | Apparatus and method for eliminating noise in stereo image |
KR101862404B1 (ko) * | 2011-12-09 | 2018-05-29 | 엘지이노텍 주식회사 | 스테레오 영상의 노이즈 제거장치 및 방법 |
JP2017005380A (ja) * | 2015-06-05 | 2017-01-05 | キヤノン株式会社 | 制御装置、撮像装置、制御方法、プログラム、および、記憶媒体 |
CN110223355A (zh) * | 2019-05-15 | 2019-09-10 | 大连理工大学 | 一种基于双重极线约束的特征标志点匹配方法 |
CN110969659A (zh) * | 2019-10-31 | 2020-04-07 | 浙江未来技术研究院(嘉兴) | 一种被动式标记点的空间定位装置和方法 |
CN110969659B (zh) * | 2019-10-31 | 2024-03-15 | 浙江清华长三角研究院 | 一种被动式标记点的空间定位装置和方法 |
Also Published As
Publication number | Publication date |
---|---|
JP5679978B2 (ja) | 2015-03-04 |
EP2549762A1 (en) | 2013-01-23 |
EP2549762B1 (en) | 2017-05-03 |
CN102428707A (zh) | 2012-04-25 |
US8890934B2 (en) | 2014-11-18 |
JPWO2011114683A1 (ja) | 2013-06-27 |
EP2549762A4 (en) | 2014-06-25 |
US20120147139A1 (en) | 2012-06-14 |
CN102428707B (zh) | 2015-03-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5679978B2 (ja) | 立体視用画像位置合わせ装置、立体視用画像位置合わせ方法、及びそのプログラム | |
US8908011B2 (en) | Three-dimensional video creating device and three-dimensional video creating method | |
US9007442B2 (en) | Stereo image display system, stereo imaging apparatus and stereo display apparatus | |
US8274552B2 (en) | Primary and auxiliary image capture devices for image processing and related methods | |
US8760502B2 (en) | Method for improving 3 dimensional effect and reducing visual fatigue and apparatus enabling the same | |
US20110063421A1 (en) | Stereoscopic image display apparatus | |
JP5814692B2 (ja) | 撮像装置及びその制御方法、プログラム | |
KR20110124473A (ko) | 다중시점 영상을 위한 3차원 영상 생성 장치 및 방법 | |
JP5420075B2 (ja) | 立体画像再生装置、その視差調整方法、視差調整プログラム、及び撮影装置 | |
US20130027520A1 (en) | 3d image recording device and 3d image signal processing device | |
JP5840022B2 (ja) | 立体画像処理装置、立体画像撮像装置、立体画像表示装置 | |
US20130050427A1 (en) | Method and apparatus for capturing three-dimensional image and apparatus for displaying three-dimensional image | |
JP5453552B2 (ja) | 撮影装置、方法及びプログラム | |
WO2012014695A1 (ja) | 立体撮像装置およびその撮像方法 | |
JP5689693B2 (ja) | 描画処理装置 | |
JP2006184434A (ja) | 立体視画像撮影装置および方法 | |
JP2012220603A (ja) | 3d映像信号撮影装置 | |
JP2015029215A (ja) | 立体画像処理装置 | |
JP2005072674A (ja) | 三次元画像生成装置および三次元画像生成システム | |
JP6022212B2 (ja) | 空間情報算出システム | |
JP2012147370A (ja) | 撮像装置 | |
JP2013223133A (ja) | 誘導装置、誘導方法、及び誘導プログラム | |
CA2982015A1 (en) | Method and apparatus for depth enhanced imaging | |
WO2013001839A1 (ja) | 撮像装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201180002079.6 Country of ref document: CN |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2011536681 Country of ref document: JP |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11755876 Country of ref document: EP Kind code of ref document: A1 |
|
REEP | Request for entry into the european phase |
Ref document number: 2011755876 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2011755876 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13320640 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |