WO2019186787A1 - 画像処理装置、画像処理方法、及び画像処理プログラム - Google Patents
画像処理装置、画像処理方法、及び画像処理プログラム Download PDFInfo
- Publication number
- WO2019186787A1 WO2019186787A1 PCT/JP2018/012852 JP2018012852W WO2019186787A1 WO 2019186787 A1 WO2019186787 A1 WO 2019186787A1 JP 2018012852 W JP2018012852 W JP 2018012852W WO 2019186787 A1 WO2019186787 A1 WO 2019186787A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- background
- dimensional
- unit
- real object
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 87
- 238000003672 processing method Methods 0.000 title claims description 13
- 230000000295 complement effect Effects 0.000 claims abstract description 11
- 230000002194 synthesizing effect Effects 0.000 claims abstract description 11
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 8
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 8
- 238000003384 imaging method Methods 0.000 claims description 47
- 235000004522 Pentaglottis sempervirens Nutrition 0.000 claims description 37
- 238000000034 method Methods 0.000 claims description 37
- 240000004050 Pentaglottis sempervirens Species 0.000 claims description 36
- 239000002131 composite material Substances 0.000 claims description 33
- 238000006243 chemical reaction Methods 0.000 claims description 19
- 238000000605 extraction Methods 0.000 abstract description 29
- 230000009466 transformation Effects 0.000 abstract 1
- 239000000203 mixture Substances 0.000 description 14
- 238000010586 diagram Methods 0.000 description 12
- 239000007787 solid Substances 0.000 description 9
- 239000000284 extract Substances 0.000 description 6
- 238000009434 installation Methods 0.000 description 6
- 238000012544 monitoring process Methods 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 102000004381 Complement C2 Human genes 0.000 description 1
- 108090000955 Complement C2 Proteins 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000001308 synthesis method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/10—Geometric effects
- G06T15/20—Perspective computation
- G06T15/205—Image-based rendering
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
- B60R1/00—Optical viewing arrangements; Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/04—Texture mapping
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/10—Geometric effects
- G06T15/20—Perspective computation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/10—Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/006—Mixed reality
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/174—Segmentation; Edge detection involving the use of two or more images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/97—Determining parameters from multiple pictures
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
- G06T2207/10012—Stereo images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30252—Vehicle exterior; Vicinity of vehicle
- G06T2207/30261—Obstacle
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30252—Vehicle exterior; Vicinity of vehicle
- G06T2207/30264—Parking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
Definitions
- the present invention relates to an image processing device that generates an overhead view composite image from a plurality of captured images, and an image processing method and an image processing program that are used to generate an overhead view composite image from a plurality of captured images.
- Patent Document 1 discloses a three-dimensional object image out of two bird's-eye view images whose viewpoints are converted based on a boundary position that divides a common imaging region between two captured images and a position of a three-dimensional object in the common imaging region. The technology which selects the bird's-eye view image with small distortion of the image and generates the bird's-eye view composite image using the selected bird's-eye view image is described.
- Japanese Patent No. 6239205 (for example, claim 1, FIG. 3)
- a bird's-eye view composite image is generated using a bird's-eye view image with a small distortion of the image of the three-dimensional object.
- the present invention has been made to solve the above-described problem, and an image processing apparatus capable of generating a bird's-eye synthesized image that does not give the viewer a sense of incongruity, and a bird's-eye synthesized image that does not give the viewer a sense of incongruity.
- An object of the present invention is to provide an image processing method and an image processing program that are used to perform the above processing.
- An image processing apparatus provides a foreground image portion in which each of a plurality of captured images is occupied by a real object that is a three-dimensional object that exists in a common shooting target area of the plurality of captured images, and the foreground image.
- An image dividing unit that divides the image into a background image portion other than the portion, and a reference image portion that is a part of a reference image that has been acquired in advance is pasted on the foreground image portion region to complement the background image portion.
- a background complement unit that generates a plurality of complemented background image parts, and a bird's-eye conversion that changes a viewpoint position of the plurality of complemented background image parts, and combines the background image parts that have been bird's-eye transformed
- a background image composition unit that generates a background bird's-eye view composite image
- a three-dimensional object recognition unit that recognizes the real object and obtains posture information of the real object, and the posture information
- a three-dimensional object projection and projection unit that acquires a three-dimensional virtual object corresponding to the real object, and a three-dimensional space superimposing unit that generates a three-dimensional space image by superimposing the three-dimensional virtual object on the background bird's-eye view synthesized image
- a display image output unit for generating and outputting an overhead view composite image that is an image of the three-dimensional space image as viewed from above.
- An image processing method provides a foreground image portion in which each of a plurality of captured images is occupied by a real object that is a three-dimensional object that exists in a common shooting target area of the plurality of captured images, and the foreground.
- the present invention it is possible to generate a bird's-eye synthesized image that hardly gives a sense of discomfort to a viewer from a plurality of captured images.
- FIG. 1 is a functional block diagram illustrating an image processing apparatus according to an embodiment.
- 1 is a diagram schematically illustrating a configuration example of an image processing system including an image processing device, two imaging devices, and a display device according to the present embodiment.
- It is a flowchart which shows the process which the solid-object extraction part of the image processing apparatus which concerns on embodiment performs.
- (A) And (b) is explanatory drawing which shows the example of the foreground image part extracted from each of the captured images by the solid-object extraction part, the background image part, and foreground image photography information. It is a flowchart which shows the process which the background complement part of the image processing apparatus which concerns on embodiment performs.
- (A) to (e) is explanatory drawing which shows the process which a background complement part performs. It is a flowchart which shows the process which the background image synthetic
- (A) to (c) is an explanatory diagram showing processing performed by the background image composition unit. It is a flowchart which shows the process which the solid-object recognition part of the image processing apparatus which concerns on embodiment performs. It is a flowchart which shows the process which the solid object projection projection part of the image processing apparatus which concerns on embodiment performs. It is explanatory drawing which shows the process which a solid object projection projection part performs.
- FIG. 1 is a diagram showing a hardware configuration of an image processing apparatus 10 according to an embodiment of the present invention.
- the image processing apparatus 10 is an apparatus that can perform the image processing method according to the present embodiment.
- the image processing apparatus 10 is, for example, a computer.
- the image processing apparatus 10 includes a processor 11 that is an information processing unit, a memory 12, a storage device 13, and an image input interface 14 that receives captured image data (also simply referred to as “captured image”). And a display device interface 15 for outputting display image data.
- the memory 12 and the storage device 13 are also referred to as a storage unit 16.
- the processor 11 performs various arithmetic processes and various control processes for hardware.
- the memory 12 is a main storage device.
- the memory 12 is, for example, a RAM (Random Access Memory).
- the storage device 13 is an auxiliary storage device.
- the storage device 13 is, for example, a hard disk device or an SSD (Solid State Drive).
- the image input interface 14 is a device for capturing a plurality of video signals provided from a plurality of imaging devices, that is, a plurality of captured images, into the image processing device 10.
- the display device interface 15 is a device for transmitting a display image to a display device such as a display.
- each of the imaging devices 20a and 20b has a function of capturing an image.
- Each of the imaging devices 20a and 20b is a camera device (also simply referred to as “camera”) provided with an imaging element and a lens such as a CCD (Charged-Coupled Device) or a CMOS (Complementary Metal-Oxide-Semiconductor).
- the imaging devices 20a and 20b are desirably camera devices having the same structure.
- the imaging device 20a images the first imaging target area.
- the imaging device 20b images the second imaging target area.
- the first imaging target area and the second imaging target area partially overlap and have a common imaging target area part.
- the imaging devices 20a and 20b may be connected to the image input interface 14 of the image processing device 10 by wire or may be connected wirelessly.
- the imaging devices 20a and 20b and the image input interface 14 communicate with each other via, for example, an IP (Internet Protocol) network or a coaxial cable.
- IP Internet Protocol
- the connection method and communication method between the imaging devices 20a and 20b and the image input interface 14 are not limited to a specific method.
- the image input interface 14 has a function of receiving two (that is, two screens) captured images 100a and 100b provided from the imaging devices 20a and 20b simultaneously (that is, in parallel).
- the two captured images 100 a and 100 b provided from the imaging devices 20 a and 20 b are taken into the image processing device 10 via the image input interface 14 and stored in the memory 12.
- the two captured images 100a and 100b captured by the image processing apparatus 10 are converted into two overhead image data (also simply referred to as “overhead images”) that are images having viewpoints above each of the imaging target regions, Thereafter, the two overhead images are combined.
- the conversion process for generating an overhead image is a “viewpoint conversion process”.
- viewpoint conversion processing for generating an overhead image is referred to as “overhead conversion processing”.
- the processor 11 reads and executes an image processing program stored in the memory 12 or the storage device 13, thereby performing viewpoint conversion processing and composition processing.
- Display image data (simply referred to as “display image”), which is overhead view composite image data (also simply referred to as “overhead composite image”) generated by the viewpoint conversion process and the synthesis process, is displayed on the display device interface 15 or the like. Sent to display device.
- FIG. 2 is a functional block diagram showing the image processing apparatus 10 according to the present embodiment.
- the image processing device 10 receives the captured images 100a and 100b from the imaging devices 20a and 20b, respectively, and outputs a bird's-eye synthesized image generated from the bird's-eye view image in the imaging target region as a display image.
- the image processing apparatus 10 extracts a three-dimensional object (also referred to as a “real object”) that is a real object from each of the captured images 100a and 100b, thereby converting each of the captured images 100a and 100b into foreground image partial data (“ It has a three-dimensional object extraction unit 1 as an image dividing unit that divides into foreground image portions ”and background image portion data (also referred to as“ background image portions ”).
- a three-dimensional object extraction unit 1 as an image dividing unit that divides into foreground image portions ”and background image portion data (also referred to as“ background image portions ”).
- the image processing apparatus 10 includes the captured images 100a and 100b (also referred to as “reference image data” or “reference image”) acquired in the past with respect to the region where the three-dimensional object is extracted in each of the captured images 100a and 100b.
- a background complement unit 2 for pasting a part of the background image portion is provided.
- the background image portion of the reference image is also referred to as “reference image portion data” or “reference image portion”.
- the image processing apparatus 10 also includes a background image synthesis unit 3 that synthesizes the background image portion of the captured image 100a and the background image portion of the captured image 100b.
- the image processing apparatus 10 corresponds to the three-dimensional object recognition unit 4 that recognizes a real object that is a three-dimensional object extracted as a foreground image part, and the foreground image part occupied by the extracted real object (that is, a three-dimensional object).
- a three-dimensional object projection projecting unit 5 that projects and projects the selected three-dimensional virtual object.
- the three-dimensional virtual object displays three-dimensional image data for displaying a virtual three-dimensional object stored in advance in the storage unit 16 or a virtual three-dimensional object having a size corresponding to the three-dimensional object. Is the three-dimensional image data generated.
- the image processing apparatus 10 also arranges (that is, superimposes) a three-dimensional space object on a background image portion formed in a virtual three-dimensional space by the background image composition unit 3;
- a display image output unit that outputs a bird's-eye view composite image formed by superimposing a three-dimensional virtual object on the background image portion as a display image;
- FIG. 3 is a diagram schematically illustrating a configuration example of an image processing system including the image processing apparatus 10 according to the present embodiment, the two imaging devices 20a and 20b, and the display device 30. is there.
- the three-dimensional object extraction unit 1 detects a real object 40 that is a real three-dimensional object from each of the captured images 100a and 100b, and extracts a foreground image portion that is a portion corresponding to the real object 40 in the captured image.
- Each of the images 100a and 100b is divided into a foreground image portion and a background image portion.
- the real object 40 is, for example, a person, a vehicle, a product, or the like.
- the three-dimensional object extraction unit 1 detects the real object 40, sets the detected real object 40 as a foreground image portion, and sets a portion other than the foreground image portion as a background image portion, thereby making each of the captured images 100a and 100b a foreground.
- the background image portion of the captured image 100a is an image portion obtained by removing the area of the real object 40 that is a three-dimensional object from the captured image 100a.
- the background image portion of the captured image 100b is an image portion obtained by removing the area of the real object 40 that is a three-dimensional object from the captured image 100b.
- the background complementing unit 2 extracts and extracts the foreground image portion that is the area of the real object 40 from the reference image stored in the storage unit 16 as a past captured image (for example, an image captured by the same imaging device).
- a past captured image for example, an image captured by the same imaging device.
- the background image portion lacking the foreground image portion is complemented.
- a background image portion is generated in which the area of the real object 40 is complemented by a part of the reference image (that is, the image data of the lacking portion is supplemented by the reference image portion data).
- the background image composition unit 3 generates a background overhead view composite image 302 from the two background image parts complemented by the background complement unit 2.
- each of the imaging devices 20a and 20b is calibrated in advance, and is acquired by the internal parameters, the external parameters, and the image processing device 10 of each of the imaging devices 20a and 20b. It is assumed that it has been.
- the internal parameters include information indicating the focal length, the position and direction of the optical axis center of each of the imaging devices 20a and 20b.
- the external parameter is information indicating the camera position and orientation, which is the position and orientation of each of the imaging devices 20a and 20b, and installation position (installation coordinate) information and installation orientation information (for example, yaw, roll, and the like) in the space to be imaged. Pitch information).
- the background image synthesizing unit 3 uses the two background image parts complemented by the background complementing unit 2 and a look-up conversion process using a reference table including pixel data indicating the correspondence between the two background image parts and the overhead view synthesized image. And the synthesis process. The processing performed by the background image synthesis unit 3 will be described in detail with reference to FIGS. 8 and 9A to 9C described later.
- the three-dimensional object recognition unit 4 first recognizes the real object 40 that is a three-dimensional object from the foreground image portion extracted from the captured image 100a and the foreground image portion extracted from the captured image 100b.
- the real object 40 is, for example, a person, a vehicle, a product, or the like. However, the real object 40 is not limited to a person, a vehicle, or a product.
- the three-dimensional object recognition unit 4 acquires posture information of the real object 40 from the foreground image portion extracted from the captured image 100a and the foreground image portion extracted from the captured image 100b, and identifies the real object 40.
- the posture information of the real object 40 is obtained by converting pixel data of two-dimensional coordinates, which are foreground image portions extracted from the captured image 100a and foreground image portions extracted from the captured image 100b, into pixel data of three-dimensional coordinates. It is a table used at the time.
- the posture information of the real object 40 may be obtained by image analysis in the foreground image portion, or may be obtained using a sensor that is a device different from the imaging devices 20a and 20b.
- the acquisition method of the posture information of the real object 40 is not limited to a specific method. The process performed by the three-dimensional object recognition unit 4 will be described in detail with reference to FIG.
- the three-dimensional object projection projection unit 5 acquires a three-dimensional virtual object 400 corresponding to the real object 40 recognized by the three-dimensional object recognition unit 4.
- the three-dimensional virtual object 400 may be selected according to the real object 40 from a plurality of three-dimensional virtual objects stored in the storage unit 16 in advance, or generated using posture information. It may be. For example, when the real object 40 is a person, a three-dimensional virtual object having a shape representing a person is used. When the real object 40 is an animal, a three-dimensional virtual object having a shape representing an animal is used.
- the three-dimensional object projection projection unit 5 performs projection projection on the three-dimensional virtual object corresponding to the foreground image portion extracted from the captured image 100a and the foreground image portion extracted from the captured image 100b by the three-dimensional object extraction unit 1.
- a projected three-dimensional virtual object 400 is generated.
- the three-dimensional object projection projection unit 5 has the shape of the three-dimensional virtual object 400 having a shape corresponding to the shape of the person seen from above at the position where the person as the real object 40 is extracted in the background overhead view composite image 302. Display the image superimposed.
- the processing performed by the three-dimensional object projection projection unit 5 will be described in detail with reference to FIGS. 11 and 12 described later.
- the processing performed by the three-dimensional space superimposing unit 6 will be described in detail with reference to FIGS. 13 and 14 described later.
- the display image output unit 7 outputs, to the display device 30, a three-dimensional overhead view composite image in which the three-dimensional virtual object 400 is superimposed on the background overhead view composite image 302 as a display image.
- the process performed by the display image output unit 7 will be described in detail with reference to FIG.
- FIG. 4 is a flowchart illustrating processing performed by the three-dimensional object extraction unit 1 of the image processing apparatus 10.
- FIG. 5A is an explanatory diagram illustrating an example of the foreground image portions 200a and 201a, the background image portion 300a, and the foreground image shooting information 500a and 501a extracted from the captured image 100a by the three-dimensional object extraction unit 1.
- FIG. 5B is an explanatory diagram illustrating examples of the foreground image portions 200b and 201b, the background image portion 300b, and the foreground image shooting information 500b and 501b extracted from the captured image 100b by the three-dimensional object extraction unit 1.
- FIGS. 5A and 5B show an example in which two foreground image portions and two foreground image shooting information are extracted from one captured image, but the number of foreground image portions is two. The number of foreground image shooting information is not limited to two.
- the foreground image shooting information includes, for example, the position coordinates of the portion of the real object 40 closest to the imaging devices 20a and 20b, the resolution of the foreground image portion, the size of the real object 40, and the like.
- the size of the real object 40 is represented, for example, by the coordinates of the four vertices of the rectangle when the real object 40 is enclosed in a rectangle (for example, circumscribed).
- the information indicating the size of the real object 40 may be an information index other than the coordinates of the four vertices of the rectangle.
- the three-dimensional object extraction unit 1 acquires a plurality of captured images 100a and 100b (step S10).
- the three-dimensional object extraction unit 1 acquires RAW image data corresponding to the captured images 100a and 100b by decoding the captured images 100a and 100b.
- RAW image data For example, from the imaging devices 20a and 20b, H.C.
- the three-dimensional object extraction unit 1 performs H.264 on the captured images 100a and 100b.
- RGBA Red Green Blue Alpha
- the format of the image data acquired by the three-dimensional object extraction unit 1 is not limited to RGBA 32-bit RAW image data.
- the three-dimensional object extraction unit 1 detects one or more real objects 40 that are three-dimensional objects such as a person, a vehicle, and a product from the acquired RAW image data (step S11).
- the real object 40 is, for example, a person walking, a traveling vehicle, a product on a production line of a factory, or the like.
- the real object 40 is not limited thereto, and may be another three-dimensional object such as an animal, a building, an obstacle, a factory facility, or a robot.
- the three-dimensional object extraction unit 1 extracts the detected real object 40 from the RAW image data, and extracts the RAW image data from the foreground image portion that is a region portion where the real object 40 is imaged and other region portions. Is divided into background image portions (step S12).
- the real object 40 is extracted by using, for example, an image segmentation technique for extracting a region of an image called a graph cut.
- the three-dimensional object extraction unit 1 can divide the background image portion and the foreground image portion from each of the captured images 100a and 100b by using the graph cut.
- the method of extracting the real object 40 is not limited to the method using the graph cut.
- a learning-based image segmentation technique using deep learning may be used to extract a real object.
- the foreground image shooting information related to the real object 40 includes, for example, the position coordinates of the foreground image portion in the captured image, a value representing the size of the foreground image portion, and an identifier for identifying the real object 40.
- Zero or one or more target real objects 40 are extracted from the captured images 100a and 100b. For this reason, the processing from the detection of the real object 40 to the extraction of the real object 40 (steps S11 and S12) is repeated the same number of times as the number of captured images to be processed.
- the three-dimensional object extraction unit 1 identifies the real object 40 with respect to the extracted real object 40 (step S13).
- the captured images 100a and 100b may capture the same real object 40.
- the three-dimensional object extraction unit 1 assigns an identifier for identifying each of the plurality of real objects to each of the real objects.
- the three-dimensional object extraction unit 1 assigns the same identifier to the real object. For example, as illustrated in FIGS.
- the three-dimensional object extraction unit 1 detects four foreground image portions 200a, 201a, 200b, and 201b, and detects the foreground image portion 201a and the foreground image portion 201b. Are determined to be the same, it is determined that the actual number of real objects is three.
- the three-dimensional object extraction unit 1 receives the captured images 100a and 100b as inputs, and outputs background image portions 300a and 300b, foreground image portions 200a, 201a, 200b and 201b, and foreground image shooting information 500a, 501a, 500b and 501b.
- FIG. 6 is a flowchart illustrating processing performed by the background complementing unit 2 of the image processing apparatus 10.
- the background complementing unit 2 performs background complementing using a reference image stored in advance in the storage unit 16 (step S20).
- the background complementing unit 2 performs background complementation using the foreground image shooting information 500a, 501a, 500b, and 501b corresponding to the target foreground image portions 200a, 201a, 200b, and 201b.
- the background complementing unit 2 calculates the position coordinates and sizes of the foreground image portions 200a, 201a, 200b, and 201b from the reference image.
- a reference image portion having the same position coordinates and size is acquired, and the reference image portion is pasted on the background image portion, thereby complementing the omission of the foreground image portion and generating the complemented background image portion.
- FIG. 7 (a) to 7 (e) are explanatory diagrams showing processing performed by the background complementing unit 2.
- the background complementing unit 2 is obtained by removing the foreground image portion 200a that is the target shown in FIG. 7B extracted from the captured image 100a shown in FIG.
- the background complementing unit 2 uses the foreground image shooting information 500a regarding the foreground image portion 200a, based on the position coordinates and the size of the foreground image portion 200a of the real object as the object, as shown in FIG.
- a reference image portion 350a having the same position coordinates and size as the position coordinates and size of the foreground image portion 200a of the real object as the object is acquired.
- the background complementing unit 2 complements the background image part 300a by pasting the reference image part 350a to the background image part 300a, and generates a complemented background image part 301a as shown in FIG. To do. That is, the background complementing unit 2 receives the background image part 300a from which the foreground image part 200a has been removed as an input, and outputs the background image part 301a that has been background supplemented using the reference image 350.
- FIG. 8 is a flowchart showing processing performed by the background image composition unit 3 of the image processing apparatus 10.
- the background image synthesizing unit 3 receives the background image portions 301a and 301b subjected to background complementing in the background complementing unit 2 as input, performs a bird's-eye conversion (viewpoint conversion) on the background image portions 301a and 301b, and the background image subjected to the bird's-eye conversion
- a background overhead view composite image 302 is generated by combining the portions.
- the background image synthesis unit 3 corrects the distortion caused by the lens characteristics of the imaging device 20a and the distortion caused by the lens characteristics of the imaging device 20b with respect to the background image portions 301a and 301b subjected to background complementation. Distortion correction processing is performed (step S30).
- the background image composition unit 3 uses the external parameters of the imaging device 20a to convert the viewpoint position so that the background image portion 301a subjected to background complementation is viewed from above (for example, from directly above). Conversion is performed (step S31). In addition, the background image composition unit 3 uses the external parameters of the imaging device 20b to convert the viewpoint position so that the background image portion 301b subjected to background complementation is viewed from above (for example, from directly above). Is performed (step S31).
- the background image synthesis unit 3 synthesizes the background image portions 301a and 301b after the overhead conversion (step S32).
- Alpha blend is an image synthesis method in which two images are superimposed and synthesized based on transparency ( ⁇ value) that is a coefficient set for each pixel.
- the ⁇ value conceptually represents the transparency from a completely opaque state having a transparency of 0% to a completely transparent state having a transparency of 100%.
- the ⁇ value is a coefficient that takes a value in the range from 0 to 1, with the minimum value (value 0) having the maximum transparency and the maximum value (value 1) having the maximum opacity. (Filled).
- FIG. 9 (a) to 9 (c) are explanatory diagrams showing processing performed by the background image composition unit 3.
- the background image composition unit 3 generates a background image portion 301a and a background image portion 301b after overhead conversion shown in FIG. 9B from the background image portions 301a and 301b shown in FIG. 9A, and A background overhead view composite image 302 shown in FIG. 9C is generated.
- the image processing apparatus 10 needs to calibrate the imaging devices 20a and 20b in advance and acquire internal parameters and external parameters.
- the internal parameters include information such as the focal length of the optical member of the imaging device, the position and direction of the optical axis center, and the like.
- the external parameter includes information on the camera position and orientation, and includes installation position (installation coordinate) information and installation orientation (yaw, roll, pitch information) in the space to be imaged.
- the background image synthesizing unit 3 can also use a reference table prepared in advance to create the background overhead image 302 from the background image portion 301a and the background image portion 301b.
- FIG. 10 is a flowchart illustrating processing performed by the three-dimensional object recognition unit 4 of the image processing apparatus 10.
- the three-dimensional object recognition unit 4 recognizes the real object 40 that is a three-dimensional object from the foreground image portions 200a, 201a, 200b, and 201b extracted by the three-dimensional object extraction unit 1 (step S40).
- the three-dimensional object recognition unit 4 acquires the posture information of the foreground image portions 200a and 200b extracted by the three-dimensional object extraction unit 1, that is, the posture information of the real object, and obtains the real object ID, the real object type, and the posture information. Is stored in the storage unit 16 (step S41).
- the posture information is a data table for converting the pixel data of the two-dimensional coordinates that are the foreground image portions 200a and 200b into the pixel data of the three-dimensional coordinates.
- the three-dimensional object recognition unit 4 may obtain the posture information in advance by image analysis in the foreground image portion, or may obtain in advance using a sensor other than the imaging device.
- the method for acquiring the posture information is not limited to a specific method. In particular, when the real object 40 is a person, since the skeleton information of the person can be acquired from the captured image, the three-dimensional object recognition unit 4 may store the skeleton information of the person in the storage unit 16 as posture information.
- FIG. 11 is a flowchart illustrating processing performed by the three-dimensional object projection projection unit 5 of the image processing apparatus 10.
- the three-dimensional object projection projection unit 5 generates a three-dimensional virtual object from the posture information of the real object acquired by the three-dimensional object recognition unit 4 (step S50).
- the three-dimensional object projection projection unit 5 projects and projects the two-dimensional foreground image portion extracted by the three-dimensional object extraction unit 1 onto a three-dimensional virtual object using posture information (step S51). Foreground image portions having the same real object ID are projected and projected onto the same three-dimensional virtual object.
- FIG. 12 is an explanatory diagram showing processing performed by the three-dimensional object projection projection unit 5.
- the three-dimensional object projection projection unit 5 acquires (including generation) the three-dimensional virtual object 400a corresponding to the recognized real object 40.
- the 3D virtual object 400a is selected from a plurality of 3D virtual object candidates stored in advance in the storage unit 16 in accordance with the corresponding real object 40. Further, the three-dimensional object projection projection unit 5 may create the three-dimensional virtual object 400a using the posture information.
- the three-dimensional object projection projection unit 5 projects the foreground image portions 200a and 200b extracted by the three-dimensional object extraction unit 1 onto the three-dimensional virtual object 400a. At this time, the three-dimensional object projection projection unit 5 performs projection projection on the three-dimensional virtual object using the posture information of the foreground image portions 200a and 200b, and generates a projected three-dimensional virtual object 400.
- FIG. 13 is a flowchart illustrating processing performed by the three-dimensional space superimposing unit 6 of the image processing apparatus 10.
- FIG. 14 is an explanatory diagram illustrating processing performed by the three-dimensional object projection projection unit 5.
- the composite image 302 is arranged (step S60).
- the three-dimensional space superimposing unit 6 arranges the projected three-dimensional virtual object 400 generated by the three-dimensional object projection projecting unit 5 so as to overlap the background overhead image composite image 302 (step S61).
- the arrangement position of the three-dimensional virtual object 400 is coordinates obtained by coordinate-converting position information included in the foreground image shooting information using internal parameters and external parameters in the imaging devices 20a and 20b.
- FIG. 15 is a flowchart illustrating processing performed by the display image output unit 7 of the image processing apparatus 10.
- the display image output unit 7 is a bird's-eye synthesized image composed of the background bird's-eye synthesized image 302 and the three-dimensional virtual object 400 arranged in the three-dimensional space generated by the three-dimensional space superimposing unit 6, that is, a designated viewpoint position ( For example, a bird's-eye view synthesized image viewed from the viewpoint position directly above the three-dimensional virtual object 400 is acquired (step S70).
- the display image output unit 7 outputs the acquired overhead view composite image to the display device 30 (step S71).
- the planar background overhead composite image 302 and the three-dimensional virtual object 400 are arranged in the three-dimensional space. Therefore, when combining the plurality of captured images 100a and 100b, the three-dimensional object is not displayed twice in the range where the captured images 100a and 100b overlap, and the three-dimensional object may disappear. Absent.
- the image processing device 10 and the image processing method according to the present embodiment when there are a plurality of real objects 40, distortion can be individually suppressed using a three-dimensional virtual object for each real object. Therefore, it is possible to generate a bird's-eye view image viewed from directly above without any sense of incongruity.
- the image processing apparatus 10 and the image processing method according to the present embodiment it is possible to create not only an overhead view viewed from directly above, but also an overhead view composite image viewed from an arbitrary viewpoint position. Therefore, when the image processing apparatus 10 is used for monitoring purposes, it is possible to improve the efficiency of monitoring work by the supervisor.
- ⁇ 4 Description of Form of Use
- the image processing apparatus 10 and the image processing method according to the present embodiment can be applied to a work monitoring system for monitoring a factory worker.
- the image processing apparatus 10 and the image processing method according to the present embodiment can be applied to a driving support system that detects and displays obstacles around the vehicle by being mounted on the vehicle.
- the image processing apparatus 10 and the image processing method according to the present embodiment can be applied to a manufacturing management system that manages work objects on a production line in a factory, an inventory management system that monitors the inventory status of finished products, and the like. .
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Graphics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Signal Processing (AREA)
- Mechanical Engineering (AREA)
- Image Processing (AREA)
- Studio Devices (AREA)
- Processing Or Creating Images (AREA)
Abstract
Description
《1-1》ハードウェア構成
図1は、本発明の実施の形態に係る画像処理装置10のハードウェア構成を示す図である。画像処理装置10は、本実施の形態に係る画像処理方法を実施することができる装置である。画像処理装置10は、例えば、コンピュータである。図1に示されるように、画像処理装置10は、情報処理部であるプロセッサ11と、メモリ12と、記憶装置13と、撮像画像データ(単に「撮像画像」とも言う)を受け取る画像入力インタフェース14と、表示画像データを出力する表示機器インタフェース15とを有する。メモリ12と記憶装置13とは、記憶部16とも称される。
図2は、本実施の形態に係る画像処理装置10を示す機能ブロック図である。画像処理装置10は、撮像装置20a及び20bから撮像画像100a及び100bをそれぞれ受け取り、撮像対象領域における俯瞰画像から生成された俯瞰合成画像を表示画像として出力する。画像処理装置10は、撮像画像100a及び100bの各々から実在する対象物である立体物(「実在オブジェクト」とも言う)を抽出することによって、撮像画像100a及び100bの各々を前景画像部分データ(「前景画像部分」とも言う)と背景画像部分データ(「背景画像部分」とも言う)とに分割する画像分割部としての立体物抽出部1を有する。
図3は、本実施の形態に係る画像処理装置10と2台の撮像装置20a及び20bと表示機器30とを含む画像処理システムの構成例を概略的に示す図である。
《2-1》立体物抽出部1
図4は、画像処理装置10の立体物抽出部1が行う処理を示すフローチャートである。図5(a)は、立体物抽出部1によって撮像画像100aから抽出された前景画像部分200a及び201a、背景画像部分300a、並びに前景画像撮影情報500a及び501aの例を示す説明図である。図5(b)は、立体物抽出部1によって撮像画像100bから抽出された前景画像部分200b及び201b、背景画像部分300b、並びに前景画像撮影情報500b及び501bの例を示す説明図である。図5(a)及び(b)には、1つの撮像画像から2つの前景画像部分と、2つの前景画像撮影情報とが抽出された例を示しているが、前景画像部分の数は2つに限定されず、前景画像撮影情報の数も2つに限定されない。
図6は、画像処理装置10の背景補完部2が行う処理を示すフローチャートである。背景補完部2は、予め記憶部16に記憶されている参照画像を用いて背景補完を行う(ステップS20)。背景補完部2は、対象となる前景画像部分200a、201a、200b及び201bに対応する前景画像撮影情報500a、501a、500b及び501bを用いて、背景補完を行う。背景補完部2は、実在オブジェクト40の前景画像部分200a、201a、200b及び201bの位置座標及び大きさを基に、参照画像から、前景画像部分200a、201a、200b及び201bの位置座標及び大きさと同じ位置座標及び大きさの参照画像部分を取得し、この参照画像部分を背景画像部分に貼りつけることで、前景画像部分の抜けを補完して、補完された背景画像部分を生成する。
図8は、画像処理装置10の背景画像合成部3が行う処理を示すフローチャートである。背景画像合成部3は、背景補完部2における背景補完が行われた背景画像部分301a及び301bを入力として受け取り、背景画像部分301a及び301bを俯瞰変換(視点変換)し、俯瞰変換された背景画像部分を合成することで背景俯瞰合成画像302を生成する。
図10は、画像処理装置10の立体物認識部4が行う処理を示すフローチャートである。立体物認識部4は、立体物抽出部1で抽出した前景画像部分200a、201a、200b及び201bから実在する立体物である実在オブジェクト40の認識を行う(ステップS40)。
図11は、画像処理装置10の立体物射影投影部5が行う処理を示すフローチャートである。立体物射影投影部5は、立体物認識部4で取得した実在オブジェクトの姿勢情報から3次元仮想オブジェクトを生成する(ステップS50)。
図13は、画像処理装置10の3次元空間重畳部6が行う処理を示すフローチャートである。図14は、立体物射影投影部5が行う処理を示す説明図である。3次元空間重畳部6は、XYZ直交座標系で示される3次元空間上に、例えば、高さ0(Z=0)の平面(例えば、XY面)に背景画像合成部3で生成した背景俯瞰合成画像302を配置する(ステップS60)。
図15は、画像処理装置10の表示画像出力部7が行う処理を示すフローチャートである。表示画像出力部7は、3次元空間重畳部6で生成された3次元空間に配置された背景俯瞰合成画像302と3次元仮想オブジェクト400とからなる俯瞰合成画像、すなわち、指定された視点位置(例えば、3次元仮想オブジェクト400の真上の視点位置)から見た俯瞰合成画像を取得する(ステップS70)。
以上に説明したように、本実施の形態に係る画像処理装置10及び画像処理方法によれば、3次元空間上に平面の背景俯瞰合成画像302と3次元仮想オブジェクト400とを配置するようにしているので、複数の撮像画像100a及び100bを合成する場合、撮像画像100a及び100bが重なる範囲において立体物が2重に表示されることはなく、また、立体物が消失することもない。
本実施の形態に係る画像処理装置10及び画像処理方法は、工場の作業者の監視用の作業監視システムに適用できる。
Claims (9)
- 複数の撮像画像の各々を、前記複数の撮像画像の共通の撮影対象領域内に実在する立体物である実在オブジェクトが占める前景画像部分と前記前景画像部分以外の背景画像部分とに分割する画像分割部と、
予め取得されている参照画像の一部である参照画像部分を前記前景画像部分の領域に貼り付けることによって前記背景画像部分を補完して、複数の補完された背景画像部分を生成する背景補完部と、
前記複数の補完された背景画像部分の視点位置を変更する俯瞰変換を行い、俯瞰変換された前記背景画像部分を合成することによって背景俯瞰合成画像を生成する背景画像合成部と、
前記実在オブジェクトを認識し、前記実在オブジェクトの姿勢情報を取得する立体物認識部と、
前記姿勢情報を用いて、前記実在オブジェクトに対応する3次元仮想オブジェクトを取得する立体物射影投影部と、
前記背景俯瞰合成画像に前記3次元仮想オブジェクトを重畳して3次元空間画像を生成する3次元空間重畳部と、
前記3次元空間画像を上から見た画像である俯瞰合成画像を生成して出力する表示画像出力部と、
を有することを特徴とする画像処理装置。 - 前記立体物射影投影部は、前記実在オブジェクトに対応する3次元仮想オブジェクトを取得し、前記3次元仮想オブジェクトに前記前景画像部分を射影投影し、射影投影された前記3次元仮想オブジェクトを前記背景俯瞰合成画像の上に重ねることを特徴とする請求項1に記載の画像処理装置。
- 複数の3次元仮想オブジェクトの候補を予め記憶する記憶部をさらに有することを特徴とする請求項1又は2に記載の画像処理装置。
- 前記参照画像は、前記複数の撮像画像を撮影した複数の撮像装置によって過去に撮影された撮像画像であることを特徴とする請求項1から3のいずれか1項に記載の画像処理装置。
- 前記参照画像を予め記憶する記憶部をさらに有することを特徴とする請求項1から4のいずれか1項に記載の画像処理装置。
- 前記立体物認識部が前記実在オブジェクトが人物であると認識した場合に、前記姿勢情報は、前記人物の骨格情報を含むことを特徴とする請求項1から5のいずれか1項に記載の画像処理装置。
- 前記表示画像出力部は、前記俯瞰合成画像として、前記実在オブジェクトを真上から見た画像を生成することを特徴とする請求項1から6のいずれか1項に記載の画像処理装置。
- 複数の撮像画像の各々を、前記複数の撮像画像の共通の撮影対象領域内に実在する立体物である実在オブジェクトが占める前景画像部分と前記前景画像部分以外の背景画像部分とに分割するステップと、
予め取得されている参照画像の一部である参照画像部分を前記前景画像部分の領域に貼り付けることによって前記背景画像部分を補完して、複数の補完された背景画像部分を生成するステップと、
前記複数の補完された背景画像部分の視点位置を変更する俯瞰変換を行い、俯瞰変換された前記背景画像部分を合成することによって背景俯瞰合成画像を生成するステップと、
前記実在オブジェクトを認識し、前記実在オブジェクトの姿勢情報を取得するステップと、
前記姿勢情報を用いて、前記実在オブジェクトに対応する3次元仮想オブジェクトを取得するステップと、
前記背景俯瞰合成画像に前記3次元仮想オブジェクトを重畳して3次元空間画像を生成するステップと、
前記3次元空間画像を上から見た画像である俯瞰合成画像を生成して出力するステップと、
を有することを特徴とする画像処理方法。 - 複数の撮像画像の各々を、前記複数の撮像画像の共通の撮影対象領域内に実在する立体物である実在オブジェクトが占める前景画像部分と前記前景画像部分以外の背景画像部分とに分割する処理と、
予め取得されている参照画像の一部である参照画像部分を前記前景画像部分の領域に貼り付けることによって前記背景画像部分を補完して、複数の補完された背景画像部分を生成する処理と、
前記複数の補完された背景画像部分の視点位置を変更する俯瞰変換を行い、俯瞰変換された前記背景画像部分を合成することによって背景俯瞰合成画像を生成する処理と、
前記実在オブジェクトを認識し、前記実在オブジェクトの姿勢情報を取得する処理と、
前記姿勢情報を用いて、前記実在オブジェクトに対応する3次元仮想オブジェクトを取得する処理と、
前記背景俯瞰合成画像に前記3次元仮想オブジェクトを重畳して3次元空間画像を生成する処理と、
前記3次元空間画像を上から見た画像である俯瞰合成画像を生成して出力する処理と、
をコンピュータに実行させることを特徴とする画像処理プログラム。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201880091553.9A CN111886624A (zh) | 2018-03-28 | 2018-03-28 | 图像处理装置、图像处理方法及图像处理程序 |
JP2018540890A JP6513300B1 (ja) | 2018-03-28 | 2018-03-28 | 画像処理装置、画像処理方法、及び画像処理プログラム |
GB2014492.9A GB2586712B (en) | 2018-03-28 | 2018-03-28 | Image processing device, image processing method, and image processing program |
PCT/JP2018/012852 WO2019186787A1 (ja) | 2018-03-28 | 2018-03-28 | 画像処理装置、画像処理方法、及び画像処理プログラム |
US17/028,508 US11403742B2 (en) | 2018-03-28 | 2020-09-22 | Image processing device, image processing method, and recording medium for generating bird's eye synthetic image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2018/012852 WO2019186787A1 (ja) | 2018-03-28 | 2018-03-28 | 画像処理装置、画像処理方法、及び画像処理プログラム |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US17/028,508 Continuation US11403742B2 (en) | 2018-03-28 | 2020-09-22 | Image processing device, image processing method, and recording medium for generating bird's eye synthetic image |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2019186787A1 true WO2019186787A1 (ja) | 2019-10-03 |
Family
ID=66530751
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2018/012852 WO2019186787A1 (ja) | 2018-03-28 | 2018-03-28 | 画像処理装置、画像処理方法、及び画像処理プログラム |
Country Status (5)
Country | Link |
---|---|
US (1) | US11403742B2 (ja) |
JP (1) | JP6513300B1 (ja) |
CN (1) | CN111886624A (ja) |
GB (1) | GB2586712B (ja) |
WO (1) | WO2019186787A1 (ja) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113538318A (zh) * | 2021-08-24 | 2021-10-22 | 北京奇艺世纪科技有限公司 | 图像处理方法、装置、终端设备以及可读存储介质 |
WO2022181174A1 (ja) * | 2021-02-24 | 2022-09-01 | ソニーグループ株式会社 | 画像処理装置、画像処理方法、プロジェクタ装置 |
WO2024142361A1 (ja) * | 2022-12-28 | 2024-07-04 | 三菱電機株式会社 | 映像合成装置、映像合成方法、及び映像合成プログラム |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019198200A1 (ja) * | 2018-04-12 | 2019-10-17 | 日本電気株式会社 | 学習用画像生成装置、学習用画像生成方法及びプログラム |
WO2021040075A1 (ko) * | 2019-08-27 | 2021-03-04 | 엘지전자 주식회사 | 영상표시장치 및 이의 영상처리방법 |
EP3968274A1 (en) * | 2020-09-14 | 2022-03-16 | Tata Consultancy Services Limited | Method and system for asset inspection using unmanned aerial vehicles |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000184336A (ja) * | 1998-10-09 | 2000-06-30 | Canon Inc | 画像処理装置、画像処理方法及びコンピュ―タ読み取り可能な記憶媒体 |
WO2012046392A1 (ja) * | 2010-10-08 | 2012-04-12 | パナソニック株式会社 | 姿勢推定装置及び姿勢推定方法 |
JP2012147149A (ja) * | 2011-01-11 | 2012-08-02 | Aisin Seiki Co Ltd | 画像生成装置 |
Family Cites Families (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050031169A1 (en) * | 2003-08-09 | 2005-02-10 | Alan Shulman | Birds eye view virtual imaging for real time composited wide field of view |
JP4934308B2 (ja) | 2005-10-17 | 2012-05-16 | 三洋電機株式会社 | 運転支援システム |
US7728879B2 (en) * | 2006-08-21 | 2010-06-01 | Sanyo Electric Co., Ltd. | Image processor and visual field support device |
JPWO2009016925A1 (ja) * | 2007-07-31 | 2010-10-14 | 株式会社豊田自動織機 | 駐車支援装置、駐車支援装置の車両側装置、駐車支援方法及び駐車支援プログラム |
JP5053043B2 (ja) | 2007-11-09 | 2012-10-17 | アルパイン株式会社 | 車両周辺画像生成装置および車両周辺画像の歪み補正方法 |
JP4697480B2 (ja) * | 2008-01-11 | 2011-06-08 | 日本電気株式会社 | 車線認識装置、車線認識方法および車線認識プログラム |
JP5206366B2 (ja) * | 2008-11-27 | 2013-06-12 | カシオ計算機株式会社 | 3次元データ作成装置 |
JP5067632B2 (ja) * | 2008-11-28 | 2012-11-07 | アイシン精機株式会社 | 鳥瞰画像生成装置 |
JP5178961B2 (ja) * | 2010-07-14 | 2013-04-10 | 三菱電機株式会社 | 画像合成装置 |
TWI433529B (zh) * | 2010-09-21 | 2014-04-01 | Huper Lab Co Ltd | 增強辨識3d物件的方法 |
WO2012053105A1 (ja) * | 2010-10-22 | 2012-04-26 | 日立建機株式会社 | 作業機械の周辺監視装置 |
WO2013046593A1 (ja) * | 2011-09-30 | 2013-04-04 | パナソニック株式会社 | 俯瞰画像生成装置、俯瞰画像生成方法、および俯瞰画像生成プログラム |
US8970701B2 (en) * | 2011-10-21 | 2015-03-03 | Mesa Engineering, Inc. | System and method for predicting vehicle location |
US20160379074A1 (en) * | 2015-06-25 | 2016-12-29 | Appropolis Inc. | System and a method for tracking mobile objects using cameras and tag devices |
US10417743B2 (en) | 2015-11-06 | 2019-09-17 | Mitsubishi Electric Corporation | Image processing device, image processing method and computer readable medium |
US10678257B2 (en) * | 2017-09-28 | 2020-06-09 | Nec Corporation | Generating occlusion-aware bird eye view representations of complex road scenes |
TWI657409B (zh) * | 2017-12-27 | 2019-04-21 | 財團法人工業技術研究院 | 虛擬導引圖示與真實影像之疊合裝置及其相關疊合方法 |
-
2018
- 2018-03-28 WO PCT/JP2018/012852 patent/WO2019186787A1/ja active Application Filing
- 2018-03-28 CN CN201880091553.9A patent/CN111886624A/zh active Pending
- 2018-03-28 JP JP2018540890A patent/JP6513300B1/ja active Active
- 2018-03-28 GB GB2014492.9A patent/GB2586712B/en active Active
-
2020
- 2020-09-22 US US17/028,508 patent/US11403742B2/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000184336A (ja) * | 1998-10-09 | 2000-06-30 | Canon Inc | 画像処理装置、画像処理方法及びコンピュ―タ読み取り可能な記憶媒体 |
WO2012046392A1 (ja) * | 2010-10-08 | 2012-04-12 | パナソニック株式会社 | 姿勢推定装置及び姿勢推定方法 |
JP2012147149A (ja) * | 2011-01-11 | 2012-08-02 | Aisin Seiki Co Ltd | 画像生成装置 |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2022181174A1 (ja) * | 2021-02-24 | 2022-09-01 | ソニーグループ株式会社 | 画像処理装置、画像処理方法、プロジェクタ装置 |
CN113538318A (zh) * | 2021-08-24 | 2021-10-22 | 北京奇艺世纪科技有限公司 | 图像处理方法、装置、终端设备以及可读存储介质 |
CN113538318B (zh) * | 2021-08-24 | 2023-12-15 | 北京奇艺世纪科技有限公司 | 图像处理方法、装置、终端设备以及可读存储介质 |
WO2024142361A1 (ja) * | 2022-12-28 | 2024-07-04 | 三菱電機株式会社 | 映像合成装置、映像合成方法、及び映像合成プログラム |
JP7584723B1 (ja) | 2022-12-28 | 2024-11-15 | 三菱電機株式会社 | 映像合成装置、映像合成方法、及び映像合成プログラム |
Also Published As
Publication number | Publication date |
---|---|
JP6513300B1 (ja) | 2019-05-15 |
GB2586712B (en) | 2021-12-22 |
JPWO2019186787A1 (ja) | 2020-04-30 |
GB2586712A (en) | 2021-03-03 |
CN111886624A (zh) | 2020-11-03 |
US20210004943A1 (en) | 2021-01-07 |
GB202014492D0 (en) | 2020-10-28 |
US11403742B2 (en) | 2022-08-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2019186787A1 (ja) | 画像処理装置、画像処理方法、及び画像処理プログラム | |
JP6425780B1 (ja) | 画像処理システム、画像処理装置、画像処理方法及びプログラム | |
JP6021541B2 (ja) | 画像処理装置及び方法 | |
CN106254854B (zh) | 三维图像的获得方法、装置及系统 | |
KR101538947B1 (ko) | 실감형 자유시점 영상 제공 장치 및 방법 | |
JP5845123B2 (ja) | 3次元モデル−インテグラル画像変換装置およびそのプログラム | |
US10349040B2 (en) | Storing data retrieved from different sensors for generating a 3-D image | |
JP6406853B2 (ja) | 光フィールド映像を生成する方法及び装置 | |
WO2019124248A1 (ja) | 画像処理装置、コンテンツ処理装置、コンテンツ処理システム、および画像処理方法 | |
JP5068732B2 (ja) | 3次元形状生成装置 | |
CN113348489A (zh) | 图像处理方法和装置 | |
WO2018052100A1 (ja) | 画像処理装置、画像処理方法、画像処理プログラム | |
US12148211B2 (en) | Image processing apparatus and 3D model generation method | |
JP2008217593A (ja) | 被写体領域抽出装置及び被写体領域抽出プログラム | |
CA2650834A1 (en) | Real-time capture and transformation of hemispherical video images to images in rectilinear coordinates | |
KR101801100B1 (ko) | 몰입형 콘텐츠 제작 지원용 영상 제공 장치 및 방법 | |
CN113132708B (zh) | 利用鱼眼相机获取三维场景图像的方法和装置、设备和介质 | |
CN109961395B (zh) | 深度图像的生成及显示方法、装置、系统、可读介质 | |
KR102019879B1 (ko) | 가상 카메라를 이용한 게임 내 360 vr 영상 획득 장치 및 방법 | |
KR102019880B1 (ko) | 분산 가상 카메라를 이용한 게임 내 360 vr 영상 획득 시스템 및 방법 | |
JP2012150614A (ja) | 自由視点画像生成装置 | |
JP7011728B2 (ja) | 画像データ出力装置、コンテンツ作成装置、コンテンツ再生装置、画像データ出力方法、コンテンツ作成方法、およびコンテンツ再生方法 | |
CN106331672A (zh) | 视点图像的获得方法、装置及系统 | |
WO2024014197A1 (ja) | 映像処理装置、映像処理方法およびプログラム | |
CN117485252A (zh) | 车辆的全景图像生成方法、装置、车辆及存储介质 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2018540890 Country of ref document: JP Kind code of ref document: A |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 18911845 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 202014492 Country of ref document: GB Kind code of ref document: A Free format text: PCT FILING DATE = 20180328 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 18911845 Country of ref document: EP Kind code of ref document: A1 |