EP3593528A2 - Dispositif et procédé pour représenter un modèle d'un objet dans un environnement virtuel - Google Patents
Dispositif et procédé pour représenter un modèle d'un objet dans un environnement virtuelInfo
- Publication number
- EP3593528A2 EP3593528A2 EP18764183.2A EP18764183A EP3593528A2 EP 3593528 A2 EP3593528 A2 EP 3593528A2 EP 18764183 A EP18764183 A EP 18764183A EP 3593528 A2 EP3593528 A2 EP 3593528A2
- Authority
- EP
- European Patent Office
- Prior art keywords
- image
- data stream
- image data
- virtual environment
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- 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
- H04N13/111—Transformation of image signals corresponding to virtual viewpoints, e.g. spatial image interpolation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
-
- 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
-
- 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
- H04N13/239—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
- H04N13/204—Image signal generators using stereoscopic image cameras
- H04N13/243—Image signal generators using stereoscopic image cameras using three or more 2D image sensors
-
- 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
- H04N13/271—Image signal generators wherein the generated image signals comprise depth maps or disparity maps
-
- 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
- H04N13/275—Image signal generators from 3D object models, e.g. computer-generated stereoscopic 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
- H04N13/20—Image signal generators
- H04N13/275—Image signal generators from 3D object models, e.g. computer-generated stereoscopic image signals
- H04N13/279—Image signal generators from 3D object models, e.g. computer-generated stereoscopic image signals the virtual viewpoint locations being selected by the viewers or determined by tracking
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/222—Studio circuitry; Studio devices; Studio equipment
- H04N5/262—Studio circuits, e.g. for mixing, switching-over, change of character of image, other special effects ; Cameras specially adapted for the electronic generation of special effects
- H04N5/272—Means for inserting a foreground image in a background image, i.e. inlay, outlay
-
- 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
- G06T2207/10016—Video; Image sequence
- G06T2207/10021—Stereoscopic video; Stereoscopic image sequence
-
- 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
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Definitions
- the present invention relates to the field of generating spatial images, that is, images that give a user a three-dimensional impression in a virtual environment, wherein the virtual environment may also include spatial images.
- Embodiments show a virtual conference room where one or more people are based on a current video of the person in real time, i. without a noticeable delay of the own movement with the movement of the space picture in the virtual space is recognizable for the person (s).
- the users can be given a complete three-dimensional impression of the persons who are in the same conference room as well as of the conference room itself.
- the conference room as well as the persons can not be generated by means of computer-animated representations, but rather based on real recorded videos.
- Embodiments show an apparatus for displaying a spatial image from an object to a virtual environment.
- the apparatus comprises a first image capture element configured to generate a first stereoscopic image data stream from an environment.
- a second image capture element can display a second stereo image element. generate scopic image data stream from the object.
- a computing unit is configured to receive the first and the second stereoscopic image data streams, starting from a reference point, based on the first stereoscopic image data stream to generate a spatial image of the virtual environment and the object, starting from the reference point from the second stereoscopic image data stream in the insert virtual environment.
- a display element can, starting from the reference point, display the spatial image of the object in the virtual environment, so that the impression of a 3D object in a 3D environment is conveyed to a viewer of the display element.
- the present disclosure is based on the idea of maximizing the user experience, ie the impression of a virtual space, by displaying both the environment and the objects displayed therein both stereoscopically and with the original video recording of the scene.
- a user may dive into a video world (rather than into a computer-animated world) of video space via video glasses or other suitable display, thereby enhancing the fidelity of the surfaces.
- the stereoscopic representation of the videos preserves the depth information of the environment as well as of the objects displayed therein, and thus also reproduces the spatial representation true to reality.
- the combination of the surface representation and the spatial representation thus gives the user the impression that he is actually in the recorded, real environment.
- Embodiments show the device with a third image capture element configured to generate a third stereoscopic image data stream from another object.
- the computing unit may receive the third stereoscopic image data stream and determine an orientation of the further object in the third stereoscopic image data stream.
- the arithmetic unit can also set disparities of the spatial image of the virtual environment as well as disparities of the spatial image of the object based on the orientation of the further object starting from the reference point.
- the arithmetic unit is designed to insert the object from the third stereoscopic image data stream as a spatial image in the virtual environment.
- the disparities or a depth map can also be generated or set starting from the reference point disparities or based on the depth map of the space image and the virtual environment, the spatial image of the object in the virtual environment, ie in the space image of the virtual environment, display.
- the terms disparity and depth map can also be used reciprocally.
- Embodiments show the device, wherein the arithmetic unit is designed to insert a spatial image of the further object from the third stereoscopic image data stream into the virtual environment and to set disparities of the spatial image of the further object based on an orientation of the object.
- the disparities can be set starting from another reference point.
- another display element can display the spatial image of the further object in the virtual environment, so that a viewer of the further display element is given the impression of a 3D object in a 3D environment.
- the viewer of the further display element can thus display the spatial image of the object from the second stereoscopic image data stream, whereas the viewer of the display element can see the spatial image of the further object displayed.
- Embodiments show the arithmetic unit which, when the reference point and / or the further reference point change, regenerate the spatial image of the virtual environment and the spatial image of the object based on the changed reference point. This can lead to the disparities being recalculated based on the changed (further) reference point.
- the display element can then display the newly calculated images.
- embodiments show the device, wherein the first image capture element is configured to generate the first stereoscopic image data stream in an initialization phase.
- the arithmetic unit can receive the first stereoscopic data stream, store it in a memory and, in an operating mode, generate a spatial image of the environment from the stored first stereoscopic data stream, depending on the reference point.
- the first stereoscopic image data stream can be generated (once) in an initialization phase and then stored locally on a memory.
- the first stereoscopic image data stream can be information, ie recordings of at least one, but optionally of a plurality of different positions in space.
- the reference points may be in the virtual environment. If the reference points are outside the positions, the representation can be estimated, eg. B. be interpolated.
- the virtual environment can also be updated during the specified time period.
- a stereo camera or a plurality of stereo cameras may be arranged in the real environment in order to record the first stereoscopic image data stream at regular intervals. All-round or 360 ° stereo cameras can be used for this, so that the first stereoscopic image data stream can also be recorded or updated in real time. This can be useful if external influences such.
- Embodiments further illustrate the apparatus, wherein the second image capture element is configured to generate the second stereoscopic image data stream from the object at a further time.
- the arithmetic unit may receive the second stereoscopic image data stream at a further time, determine a position and / or orientation of the object at the further time from a reference point, and based on the position and / or orientation of the object at the further time adjust the spatial image of the object in the virtual environment.
- the arithmetic unit for.
- a 3D engine that dynamically and optionally adjusts room images of both the object and the virtual environment in real time.
- the reference point can therefore be set dynamically.
- the arithmetic unit can then redetermined and set the disparities based on the new reference point and an orientation thereof.
- a person who stands on the reference point can be assumed for generating the spatial images or determining / adjusting the disparities. The person may also look in a certain direction, thereby characterizing the orientation of the reference point.
- the position and / or the orientation can calculate the arithmetic unit from the third image data stream by z. B. the position and / or orientation of the further object from the third image data stream to a receiving (stereo) camera is determined.
- a 3D view of the objects, in addition to the stereoscopic view, the one spatial depth also provides a panoramic view to allow the object and thus to include this from every page.
- Embodiments also show the arithmetic unit which can superimpose sub-picture data on at least a part of the spatial image of the object.
- the object and / or the further object if they are persons, a face or eye area of another, for example, prominent person are superimposed.
- the person wears data goggles during the production of the image data stream to acquire the goggles in the virtual environment with an image of the eye area or the face of the person without data goggles, which is kept in a memory, for example replace or overlay.
- the sub-picture data may also be recorded stereoscopically, so that the overall impression of the user by the sub-picture data is still realistic.
- the description of the object can also be applied to the further object. Only the reference point and the corresponding image acquisition elements or display elements are to be adapted accordingly.
- the arithmetic unit may determine an actual appearance of the object and adjust the sub-picture data based on the determined current appearance or select sub-picture data from a plurality of sub-picture data based on the determined current appearance and superimpose the selected sub-picture data on the part of the space picture.
- the (second or third) image capturing element can capture an eye area or a face of a person as an object and store it as overlay image data.
- the arithmetic unit can superimpose, for example, a pair of data goggles placed on the person with the overlay images obtained in the initialization phase.
- the arithmetic unit may adjust the eye area or the face in the sub-picture data based on a state of mind (here, the appearance) of the person and superimpose the adjusted sub-picture data of the person.
- z. B. in the superposition of the data glasses with the face or the eye area from the overlay image data prominent areas in the overlay image data in the face such as the eyes or the mouth can be adapted to a state of mind of the person. So can be closed by upwardly pointing corner of the mouth on a laugh of the person, so that the inserted eye area laughing, so for example slightly narrowed eyes represents.
- This modification is applied, for example, to static overlay image data by means of an image processing algorithm so that the static overlay image data is dynamic.
- a plurality of sub-picture data are stored or stored, for example, by comparing the current recording of the mouth with a mouth from the stored sub-picture data, a matching eye-area or face to be overlaid is obtained.
- the image pickup unit may apply a plurality of sub-picture data, wherein the arithmetic unit is configured to compare the state of mind based on a comparison of a current position of the mouth angle of the person with a position of the corners of the mouth in the plurality of sub-picture data and the sub-picture containing the current facial expression best possible to select.
- a method may represent the operations performed by the device.
- the method can be mapped by a program code which, when executed on a computer, performs the corresponding method steps.
- FIG. 1 shows a schematic representation of a conventional approach to the production, transmission and display of a stereoscopic film comprising a multiplicity of individual images
- FIG. a schematic representation of a procedure for the production, transmission and presentation of a stereoscopic film with single models generated from a 3D engine
- Fig. 3 shows schematically the procedure for removing the background of a
- FIG. 3 (a) shows the scene initially taken, and wherein Fig. 3 (b) shows the scene with no background;
- 5 is a flowchart of an embodiment in which the 3D model under
- FIG. 6 is a flowchart illustrating an embodiment for changing a human face
- FIG. 7 shows a flow chart of an embodiment of the described method for producing a three-dimensional representation of at least one part of an object, as is carried out, for example, by the system described with reference to FIG. 2;
- Fig. 8 is a schematic block diagram of the apparatus for displaying a
- Fig. 9 is a schematic block diagram of the apparatus for displaying a
- FIG. 8 shows a schematic block diagram of a device 800 for displaying a spatial image from an object into a virtual environment.
- the device 800 has a first image capture element 802 and a second image capture element 804.
- a computing unit 806 may receive, process and output the image data of the two image capture elements to a display element 808.
- the first image capture element 802 may generate a first stereoscopic image data stream 810 from an environment.
- the image data so z.
- a video can be used to form the virtual environment. This is for example a meeting room, it can but also to any other scenery in which (as described below) an object is inserted.
- the second image capture element may generate a second stereoscopic image data stream 812 from the object.
- the arithmetic unit 806 receives both image data streams 810, 812.
- the first and second sheet detection members may be physically different. It can therefore be different cameras. It is thus possible to record the environment spatially and temporally independently of the object, for example in an initialization phase. Also, the object may be at a different location, for example, in a room that is optimized to free the object from the background by means of color-based image chroma keying. It can therefore be a "blue screen” or "green screen” box. However, it is also possible that the object is actually in the real environment, which is then recorded, for example, in advance. In this case, the first and second image capture elements can also be the same camera, which, however, records the environment and the object at different times. Thus, the distinction between the first and second image sensing elements may also be purely functional rather than physical.
- this z. B a 360 ° stereo camera can be used, which records in all directions (or a plurality of discrete directions) a stereoscopic Operajstein- ström from a recording point.
- This recording can also be carried out from different recording points in order to be able to set disparities starting from the reference point based on real recordings.
- the reference point is at a distance from a real pick-up point, then the view can also be interpolated or calculated.
- stereo cameras can be arranged in the room in which the person is for the recording.
- the stereo cameras can be aimed at the person and record them from different sides.
- 5 or 7 or any other number of stereo cameras are used to simultaneously capture the object from different sides and to generate with each camera a stereoscopic sub-frame data stream that composes the second stereoscopic image data stream. Since the stereo cameras typically have a distance to each other, they can not provide every possible angle to the object.
- the arithmetic unit When the reference point is changed to a viewing angle which differs from a viewing angle of one of the stereo cameras, the arithmetic unit has the possibility of viewing the object from z. B. interpolate two stereo cameras have a similar angle. Furthermore, however, a so-called billboard technique can also be used. In this technique, the spatial image of the object is always rotated to the reference point so that the viewing angle of one of the stereo cameras coincides with the viewing angle of the reference point. As a corresponding stereo camera advantageously the stereo camera is selected, which comes closest to the viewpoint of the reference point.
- the arithmetic unit 806 may receive the first and second stereoscopic image data streams 810, 812. Starting from a reference point, based on the first stereoscopic image data stream, a spatial image of the virtual environment can now be generated. In addition, a spatial image of the object, starting from the reference point, can be inserted from the second stereoscopic image data stream into the virtual environment. The spatial images can be generated based on disparities that are matched to a person standing on the reference point and having a particular line of sight. The arithmetic unit can thus take over the 3D rendering, or more precisely the generation of the spatial images from the stereoscopic image data streams.
- the arithmetic unit may first select the stereoscopic sub-picture data stream of the first stereoscopic image data stream and / or the second stereoscopic image data stream that comes closest to the reference point. A possible difference between the reference point and a pick-up point from which the environment or the object was recorded can be deducted. For this purpose, among other techniques, an interpolation between two pickup points can be used.
- the display element 808, starting from the reference point, can display the spatial image of the object in the virtual environment. In this way, the impression of a 3D object in a 3D environment can be conveyed to a viewer of the pointing element.
- the display element 808 uses, for example, the position and direction for the show that the arithmetic unit has already used to generate the spatial images. In other words, the display element displays the data 814 provided by the arithmetic unit. These are the space image 814a of the object and the space image 814b of the virtual environment.
- the device 800 further includes a third image sensing element 816.
- the third image capture element can have all the features of the second image capture element 804, but a different, further object is recorded. From this further object, the third image data stream 818 is generated and transmitted to the arithmetic unit 806.
- the arithmetic unit extracts the further object from the third stereoscopic image data stream 818, generates an optional spatial image 814c from the further object and the same into the virtual environment.
- the display element also displays the virtual environment. If, taking into account the reference point and the viewing angle, the spatial image of the object or the spatial image of the further object is in the direction of view, the corresponding spatial image is also displayed.
- the image data streams may be a sequence of individual images, for example a (live or real-time) video or video stream.
- the image capture element 802 can record the environment 820 before the actual usage and transmit the first stereoscopic image data stream 812 to the computing unit 806.
- the arithmetic unit 806 may be constructed in a decentralized manner, so that a partial computing unit 806a is arranged on the side of a first user 822a and a second partial computing unit 806b is arranged on the side of a second user 822b.
- the second image capture element 804 may now record the first user 822a and the third image capture element 816 the second user 822b and generate the second and third stereoscopic image data streams 810 and 818, respectively, and transmit them to the respective partial computing unit.
- connection 824 the two stereoscopic image data streams can be exchanged between the two particle computation units.
- the connection is z. B. via the Internet.
- the first stereoscopic image data stream can also be transmitted via the Internet to the first and second partial computing units 806a, 806b.
- the first partial computing unit can now generate the spatial image of the further object as well as the spatial image of the environment and transmit both for display to the display element 808.
- the data stream 814a can therefore be the spatial image of the further observer. and the spatial image of the environment.
- a current position and / or a current viewing direction of the first user 822a can be used as a reference point for the calculation of the spatial images or for the determination of the depth map.
- a reference point in its real environment usually this is not the environment 820
- a reference point in its real environment (usually this is not the environment 820) can be specified, which is mapped to a reference point in the virtual environment.
- a deviation from this reference point in the real environment of the first user can then be provided analogously or with a corresponding scaling factor, transmitted to the virtual environment in order to set the reference point to the corresponding point in the virtual environment. Likewise, it is possible to proceed with the orientation of the user.
- the spatial image of the environment with the spatial image of the object can now be displayed.
- the object can be arranged here on the reference point determined above and have a viewing direction in accordance with the orientation.
- the position and / or orientation of the further object 822b can now be used, which can be determined analogously to the determination thereof in the object 822a.
- a computing system 826a, 826b each having an image capture element (one or more stereo cameras), a display element, for. B. a screen or a VR (virtual reality) glasses and a corresponding computer has.
- a computing system 826a, 826b each having an image capture element (one or more stereo cameras), a display element, for. B. a screen or a VR (virtual reality) glasses and a corresponding computer has.
- other topologies or subdivisions into subunits of the device are also possible.
- the second and / or the third image-capturing element can record sub-picture data in an initialization phase, ie before the actual operation of the device.
- These sub-picture data are z.
- These sub-picture data may also be recorded once for multiple use and then reused in each mode of operation.
- this z. B. find where the objects, so in this case the user, wearing (in the operating mode) VR glasses.
- the arithmetic unit can now retouch the VR glasses or overlay them with the overlay image data so that the VR glasses are no longer visible. The users thus have the impression that the other user without VR glasses is filmed, although both users wear such glasses.
- the overlay image data may be extracted from a corresponding image or image data stream in the initialization phase, and, e.g. B. as well as the image data stream 812 store in a memory of the arithmetic unit.
- the display element can display the spatial image of the object, for example a user, with the superimposed or textured overlay image data.
- the sub-picture data may be static (frame) or dynamic ([short] video sequence).
- sub-picture data are conceivable. It is thus possible, for example, to superimpose the face of a (famous) person over the face of the spatial image of the object instead of the eye area. Alternatively, it is also conceivable to cover one's own face or the eye area with a mask in order to preserve its anonymity in a chat room, for example. Such a mask can also be extracted without an explicit record, for example, from a corresponding stereographic photo or video. Optionally, a computer animation is also conceivable here, since the mask does not necessarily depend on the real impression.
- natural markers such. B. nose, ears or mouth of the object are used.
- external markers ie z. As light emitting diodes / lamps or reflectors are attached to the object.
- the arithmetic unit is configured to determine an actual appearance of the object 822a, 822b and to adjust the sub-picture data based on the determined current appearance.
- the manifestation is, for example, a smile or another facial expression of a person. This can be determined from the facial features, eg the corners of the mouth.
- Static overlay image data such as a single image, can now be manipulated such that the eye area in the overlay image data matches the current facial expression.
- the eye area can be processed in such a way that they appear slightly pinched in order to present a smile.
- dynamics in the sub-picture data can be generated.
- a plurality of sub-picture data may also be applied.
- the current position of the mouth angle of the person can be compared with the position of the corner of the mouth in the overlay image data.
- the overlay image that best reflects the current facial expression and to match the entire face or even only the eye area of the space image with the overlay image data overlap.
- a dynamic in the sub-picture data can also be generated from the static sub-picture data.
- the arithmetic unit may select sub-picture data from a plurality of sub-picture data based on the determined current appearance, and superimpose the selected sub-picture data on the part of the object.
- the image capture element in the initialization phase, can be designed to scan an eye area or a person's face as an object and to store it as overlay image data, wherein the computing unit is designed to have a person-eye-mounted data goggle with those obtained in the initialization phase in the operating mode Overlay images superimpose.
- the arithmetic unit can adjust the eye area or the face in the sub-picture data based on a mood of the person and superimpose the adjusted sub-picture data of the person.
- the image pickup unit may apply a plurality of sub-picture data, wherein the arithmetic unit is configured to compare the state of mind based on a comparison of a current position of the mouth angle of the person with a position of the corners of the mouth in the plurality of sub-picture data and the sub-picture, the best possible the current facial expression to select.
- the current appearance of the object may be determined 30 to 60 times per second, 35 to 55 times per second, or 40 to 50 times per second.
- the current appearance of the object for example the facial expression of a person, for each image or each frame of the image data stream, as an average or median of a plurality of images or frames of the image data stream is determined.
- the image data stream may also have 30 to 60 frames per second, 35 to 55 frames per second or 40 to 50 frames per second, or a multiple thereof. This comparatively high refresh rate makes it possible to update the spatial image in the virtual environment in real time, that is, assuming a sufficient transmission speed, no difference visible to the human being between the object in the real environment and the object in the virtual environment can be recognized.
- the latency of the object in the virtual environment compared to the object in the real environment is less than 50 ms, less than 30 ms, or less than 20 ms.
- a user for example by means of an input device, to set his viewing angle in the virtual environment 820 independently of his position in the virtual environment.
- the user can obtain an individual view of an external observer (third person view) on the room images of himself and the other user in the virtual environment 820. Users can virtually sit face to face and hold their meeting in the virtual conference room.
- visual objects as a model, room image or hologram in the virtual conference room.
- the model of an object for example a car, which is true to the original, can be integrated into the conference room whose orientation, size and / or level of detailing can be adapted individually and independently by each of the two persons.
- Embodiments show that both the image data stream and the overlay image data are generated or recorded by a camera.
- the user can be given an improved impression of reality than with computer-generated avatars or avatars in which, for example, only the upper body is recorded and computer-animated legs are inserted to give the avatar in the virtual world the opportunity to move.
- the recorded camera image can be converted by means of the procedure described in FIGS. 1 to 7 into a 3D spatial image in order to enable a reduction in the amount of data to be transmitted.
- the data reduction is not limited to this method and moreover it is completely optional, since the uncompressed recorded camera images can also be transmitted.
- the overlay image data can already be recorded in advance and stored in the arithmetic unit. Since one field of application of the device described is a virtual meeting or conference room, the arithmetic unit can be arranged decentrally with the participants of the conference. In order to use the overlay image data efficiently, they can be exchanged, for example, once between the particle computation units of all participants or sent to them. For this purpose, the recorded images or photos or video sequences of the sub-picture data can be used without converting them into a corresponding computer model.
- the recorded objects in the image data stream and / or the sub-picture data from their real background can be recorded, for example, in a "green box" (in front of a green or a blue background), so that the object can be read by means of color-based image release , chroma keying) can be inserted into the virtual environment.
- a green box in front of a green or a blue background
- Embodiments also show that the apparatus shown can be used to visualize cosmetic corrections.
- sub-picture data e.g. be stored differently lips so that the user from the set of stored overlay image data, the corresponding make-up lips of his choice can overlay his own lips.
- users can get a better impression of whether the chosen lipstick color matches the rest of the outfit, as they see both in combination.
- the user may once apply different lipsticks in advance and in an initialization phase and store the overlay image data as a shot from his own lips.
- the user can thus also in the virtual world and thus, for. make-up for a (professional) meeting without spending time in make-up in real life.
- the manufacturers of lipsticks corresponding pattern can be provided as overlay image data.
- the lips have been selected here by way of example, but the principle can be used for any cosmetic corrections, such as a different hairstyle or hair color. Likewise, thus also an existing reading or seeing glasses can be retouched.
- a similar principle can be applied to the general change of (facial) attributes.
- the sub-picture data may be manipulated (with the computer) to alter certain attributes of the user. This principle can e.g. used in cosmetic surgery to give the user an idea of what it looks like after a potential procedure.
- a corresponding method for displaying a spatial image from an object to a virtual environment may include the steps of: generating a first stereoscopic image data stream from an environment; Generating a second stereoscopic image data stream from the object; Receiving the first and second stereoscopic image data streams; Generating a spatial image of the virtual environment starting from a reference point based on the first stereoscopic image data Strom; Inserting the object, starting from the reference point, from the second stereoscopic image data stream into the virtual environment; and displaying, based on the reference point, the spatial image of the object in the virtual environment. Further embodiments of the device can also be applied to the method accordingly.
- a mesh model of the object may be generated from, for example, the second stereoscopic image data stream.
- the grid model is invisible to the user, but contains the depth information of the object from the perspective of the stereo camera used.
- the mesh model can then be inserted into the virtual environment and overlaid with the video images without depth information.
- the transmission of an image data stream of the stereoscopic image data stream (this contains at least two image data streams) is sufficient to obtain both the real image of the object and the corresponding depth information for generating the spatial image from the object.
- model and room image refer to the same object, the image of the object in the virtual environment.
- the spatial image already implies the generation by means of stereoscopic image data, whereas in the model a simple image data stream is sufficient.
- All embodiments which are directed to a model are analogous to a spatial image transferable.
- the model has a reduced amount of data when storing or transmitting the model.
- the aforementioned embodiments can be performed even at low transmission rates or bandwidths in real time.
- the following embodiments are only illustrative and do not limit the creation of the model to the same.
- Fig. 1 shows a schematic representation of a conventional approach for the production, transfer and presentation of a stereoscopic film comprising a plurality of individual images.
- a cube 100 is shown in FIG. 1 as an object to be displayed.
- a first camera 102 generates a first shot 100a of the cube from a first perspective
- a second camera 104 generates a second shot 100b of the cube from a second perspective that is different from the first perspective.
- the receptacles 100a and 100b of the cube 100 are generated from different angles.
- the thus generated and received frames 100a, 100b are provided to a suitable stereoscopic display unit 108, such as a monitor, for display.
- a suitable stereoscopic display unit 108 such as a monitor
- a common 3D camera can also be used, which likewise generates two images of the object 100, which are then transmitted to the monitor 108 for display in the manner described above.
- the conventional approach described with reference to FIG. 1 is disadvantageous, since the amount of data to be transmitted via the transmission medium 106 of at least two images 100a and 100b is very large, which is accompanied by a correspondingly high data transmission time. Even with a compression of the two-dimensional images or images 100a, 100b, the time required for the compression of the data is large, so that the total transfer time from the location where the image of the object 100 is generated to the location where the three-dimensional rendering should be done, is very high.
- a stereoscopic image of the cube 100 is assumed, and in the procedure described with reference to FIG.
- the just mentioned transmission of a live stream in three-dimensional quality or the transmission of a three-dimensional live sequence is desired, for example, in connection with the recording of people and surrounding spaces.
- the inclusion of humans and surrounding spaces using 3D cameras and playing back as a 3 D film has the above-mentioned problem of immense amounts of data that are not transferable on the Internet, which is due to the fact that the data is as conventional 2D sequences -Bildem must be stored and transferred.
- FIG. 2 shows a schematic representation of a procedure for producing, transmitting and displaying a stereoscopic film with individual models generated from a 3D engine. Similar to FIG. 1, the three-dimensional image of a cube 100 is also selected in FIG. 2 as the basis for explaining the approach.
- FIG. 2 shows a schematic overall representation of the system 200 according to exemplary embodiments for generating a three-dimensional representation of at least one part of an object, namely the cube 100.
- the system 100 comprises the transmission medium 106 already described with reference to FIG and a receiver side 204. Elements which have already been described with reference to FIG. 1 are provided with the same reference symbols in FIG. 2 and will not be explained again in detail.
- the transmitter side 202 shown in FIG. 2 may, for example, comprise a device for generating data for a three-dimensional representation of at least a part of an object, wherein the device according to the exemplary embodiment shown comprises a 3D engine 206, which is shown schematically in FIG. and receiving as input the SD picture comprising the two pictures 100a and 100b of the cube 100. Based on the 3-D image, the 3D engine generates a model 208 of the cube 100 comprising the vertices A to H located in space at different positions. Depending on which different perspectives the cube 100 is captured from, this includes the 3D engine 206 generated 3D model either the entire object, so the entire cube 100 or only a portion of the cube 100. In the embodiment shown in FIG. 2, it is assumed that only a single 3D image of the cube 100 exists from a perspective, so that the 3D model according to this embodiment would represent only the front and the top of the cube, and correspondingly the vertices A to G exist.
- the 3D engine 206 would be based on the input data, namely based on the two 3D images, create a 3D model of the cube 208, which includes the vertices A to H, ie a 3D model that can be viewed from all sides.
- the 3D engine 206 is operative to define the 3D model 208 of the cube 100 in the form of the vertices A through H and as color values associated with the respective vertices.
- the scene captured by the camera 102, 104 which includes, for example, the cube 100, provides at the output of the 3D engine 206 a frame including color values and depth information.
- the color values consist in each case, according to one embodiment, of one byte for red, one byte for green and one byte for blue (RGB color values), ie 3 bytes in total.
- the depth information includes the X, Y, and Z values for the vertices of the 3D model in three-dimensional space.
- the X, Y, and Z values may each be stored as single precision floating point numbers, for example, 32 bits long.
- the sequence of X, Y and Z values are referred to as vertices, and the set of all vertices of the SD model 208 is referred to as a point cloud.
- the example of the generation of data for the three-dimensional representation of a cube 100 shown there does not represent a restriction; in fact, the object 100 can be any object of arbitrarily complex structure, for example also a representation of a person or a machine.
- the 3D engines shown in Fig. 2 may be e.g. be implemented by computer systems, which are possibly equipped for a generation of 3 D models hardware technically and on which appropriate software is provided for execution.
- the device 202 is configured to repeatedly generate a 3D image 100a, 100b of the object 100 to provide data for a plurality of consecutive frames for transmission over the medium 106 so as to provide a live 3D sequence on the receiver side 204 or a 3D movie.
- up to 30 frames per second are generated by device 202, i. Up to 30 frames are taken by the object 100.
- Each frame is encoded via the 3D engine 206 using the 3D model, as described above, so that each of the 30 frames per second contains a dataset containing the vertices and the color values of the object 00 associated with the vertices at the time of acquisition ,
- the device 202 and the method implemented thereby is advantageous because an amount of data that is transmitted via the transmission medium 106 is significantly reduced, which also significantly reduces the transmission time.
- the time for calculating the 3D models is shorter than that for compressing individual images according to Time required in the prior art.
- the complete transmission of the 3D information via the 3D model on the receiver side makes it possible to freely choose a viewpoint on the object on the receiver side, since the actual images after transmission of the data of the 3D model at the receiver side by a 3D engine at runtime be generated and displayed.
- the data (the data set) describing the 3-D model 208 generated on the transmitter side is transmitted via the transmission medium 106 to the receiver side 204, so that the corresponding data representing the 3D model 208 'is transmitted. be present on the receiver side 204.
- These data are fed to a 3D engine 210 which, on the basis of the 3D model, generates the 3D image of the object in the corresponding frame, for example the two images 100a, 100b, which then, as in the prior art, a suitable monitor 108 or other suitable display device for the three-dimensional representation of the object on the receiver side 204 are provided.
- an approach is taught in which a transmission of a sequence of individual 3D models is performed, rather than a transmission of a sequence of individual 2D images, as is the case in the prior art.
- the SD models 208 are generated prior to transmission using the 3D engine 206, where the 3D engine recognizes edges from the images 100a, 100b, for example, and generates the 3D model based on the detected edges.
- the 3D engine 206 may determine common areas in the images 100a, 100b, eg, common edges associated with the object 100 to determine the resulting 3D mesh from the object 100.
- the 3-D model described by the vertex and color values is converted back into the two images 100a and 100b by the receiver-side 3D engine 210 to image the 3-D object from different angles with the individual images of a and b which are then displayed on the stereoscopic output device 108.
- the object is a person who is inside a scene.
- the scene encompasses the human being, who stands in a room and easily moves back and forth in his or her place.
- the device 202 (see FIG. 2) records 30 frames per frame of this scene per second, and generates for each frame a corresponding 3D model of the scene and describes it by the vertices and color values.
- each frame (each As mentioned above, the color values and the depth information, for example RGB values and XY and Z values, each defining a vertex, the plurality of vertexes forming a point cloud.
- it may be desirable to extract the background from the scene for example, when only the person's image is to be transmitted to the receiver side 204 (see FIG. 2) and displayed there for example with a different background, either from the scene Sender is transmitted to the receiver in advance or in parallel or is specified at the receiver.
- the Z value of each vertex is compared with a corridor distance (threshold distance) of the standing area of the person to be represented, and distances indicating that a vertex is farther away or closer than the corridor distance is recognized as a background vertex and omitted. ie, only those vertices are allowed that are within the corridor distance, and the data thus obtained are those that map the person in a point cloud. In other words, the number of vertices that define the point cloud per frame is reduced by those that are clearly assigned to the background.
- FIG. 3 schematically illustrates this approach, with FIG. 3 (a) showing the initially recorded scene 300, in which a person 302 (object) is depicted standing on a floor 304. Furthermore, a rear wall 306 is shown. Along the Z-coordinate is shown the area 308 bounded by the dashed lines 308a, 308b extending in the X-direction.
- the region 308 is the above-mentioned threshold corridor, and according to embodiments, vertex values of a 3D model generated due to the 3-D image of the scene 300 are deleted, provided that they are outside the threshold corridor 308, so that the in Fig. 3 (b) shows that only the person 302 and part of the bottom 304 'remain.
- the data model of the person 302 generated by the 3D engine may be, for example, a mesh model or a triangular mesh, depending on the 3D engine used.
- a mesh model or a triangular mesh, depending on the 3D engine used.
- the edge region of the 3D model can be smoothed by a smoothing algorithm, for example by a smoothing algorithm that filters out large depth values or depth separations.
- an object 100, 302 can be picked up from multiple perspectives.
- One reason for the multiple shots may be that the object is to be completely imaged, so that a complete 3 D model is present.
- a situation may arise that due to the design of the object, for example due to a concealment of a part the body of a person by a hand of the person or by a projecting portion of an object underlying sections are not detected by a 3 D shot.
- these holes are caused by the fact that the distance of the infrared sensor and the imaging camera is a few centimeters, so that the viewing pyramids of both sensors are not completely covered.
- perspective occlusions e.g. a hand of a person in front of their body, thus arise areas that have no triangular mesh or no portion of the 3 D model as a basis, or even areas that have no image texture.
- the textured 3D models of the person's various overlapping views obtained in the manner described above are subsequently connected, according to embodiments, to a 360 ° all-round view of the person.
- Each two overlapping triangular meshes are brought to overlap in overlapping areas and starting with a given plane, for example a horizontal plane (X-Z plane), those vertices are identified which have a predetermined distance from each other.
- a given plane for example a horizontal plane (X-Z plane)
- those vertices are identified which have a predetermined distance from each other.
- the distance can be varied. For example, depending on the later transmission of the data, with knowledge of the transmission medium, the amount of data can be adapted to a bandwidth and the distance determined accordingly.
- the identified points are grouped together in a new triangle mesh, and when a point set, for example, a circular point set is found on a plane, the method moves to the next higher level, which is repeated until the relevant point cloud or vertices for the outer shell of the object are found. For example, all the resulting points can be displayed from bottom to top in a linked spindle. As a result, a textured connected point cloud of the outer shell of the object is obtained as a frame, in short, a plurality of X, Y, Z values in an order.
- a 3D model of the part of the object is generated using a 3D image of the part of the object.
- data is provided using the 3-D model describing the vertices of the part of the object in three-dimensional space and the color values associated with the vertices.
- the method includes, as an optional step, extracting the background S104 from the data using the Z value of each As discussed above, and as a further optional step, correcting S106 of a peripheral area of the object by filtering out of depth intervals exceeding a predetermined threshold, as discussed above.
- edge region correction may include anti-aliasing and avoidance of spikes containing large steep slope depth values.
- Steps S100 through S106 may be used to generate a 3-D live sequence or 3D movie for each frame of the sequence, respectively / of the movie, each frame being a frame of the sequence / movie.
- the repetition rate is 30 frames per second.
- texture information may be interpolated between the vertex values (vertices) and thus requires little data width.
- step S108 a first 3D model is generated using the first 3D shot and a second 3D model using the second 3D shot.
- step S1 10 the first 3 D model and the second 3 D model are connected to generate a common 3 D model, using the data provided in step S102 (see FIG generated and deployed in a common 3D model.
- the joining of the two SD models according to the illustrated embodiment comprises the step S110a, in which the two 3D models are arranged so that their overlapping areas are in coincidence.
- step S1 10b the vertices that are within a predefined distance are defined, and the identified vertices are summarized in step S1 10c.
- Steps S1 10a to S1 10c are repeated if it is determined in step S1 12 that not all predetermined levels relating to the 3-D model have been processed yet. In this case, another plane is selected in step S1 14, and the process returns to step S1 10b. Otherwise, if it is determined in step S1 12 that all levels have been processed, the method ends in step 1 16.
- the object may also be the background of a Be room to be displayed on the receiver side so that a user on the receiver side can view the room from different perspectives and can move within predetermined limits within the room.
- the 3D image includes a 3D image of the background of a room, for example, according to the steps S100 to S116, but without the step S104, as removal of the background at this point is of course not meaningful.
- the steps in block S110 result in a connection of the various receptacles of the interior to produce the inner envelope of the space.
- it can be provided to define an area, for example a circular area, in which a user can "move freely” when displaying the thus generated three-dimensional space on the receiver side, so as to create the illusion of a live 3D image. To get movies.
- FIG. 6 shows a flowchart illustrating the embodiment for changing a human face.
- the method is based on an object representing the face of a person, and in a step S1 18, a static facial model of an original face of another person is first provided.
- the first position may be, for example, a so-called look-a-like person who looks similar to the other person, for example a known person.
- the face model of the other person's original face is a textured 3D model with texture that example, from a photo or film recording of the other person has emerged and thus can have a correspondingly high recognition effect.
- a position of the person's face in space is determined upon generation of the 3D shot.
- a positioning device e.g. Nexonar, a device equipped with a sonar sensor that is worn on the back of the head, or a raycasting device.
- the 3D model of the person's face is superimposed on the other person's static face model, and in step S124, the 3D model of the person's face is adjusted to those points where there is no movement, to the other person's static facial model.
- the difference between the two 3D models or triangular gratings may be added at those locations where there is no movement, e.g.
- step S126 a texture is created from the 3D image of the person's face, at those locations where there is movement, to create a hole mask texture that is semi-transparently textured to the common or new 3D model in step S128 so as to obtain, at runtime, a 3D model that is recognizable to human eyes as a sequence of moving and animated representation of the original face.
- step S130 An exemplary embodiment for generating a three-dimensional representation of at least one part of an object is explained in more detail below with reference to FIG. 7, as is carried out, for example, by the system described with reference to FIG. 2.
- step S130 data are generated for the three-dimensional representation of the at least one part of the object, in accordance with the method described above, as explained, for example, with reference to FIGS. 4, 5 and 6 or as with the receiver side 202 in Fig. 2 has been explained.
- step S132 transmission of the data via the transmission medium 106 is performed from the transmitter side 202 to the receiver side 204 (see FIG. 2), and in step S134, the 3D images 100a and 100b shown in FIG 3D engine 210 generated on receiver side 204.
- step S 36 The production of the 3D image is done by a 3D engine to stereoscopically display or project the object.
- the 3D image is displayed by a display device, for example an autostereoscopic 3D monitor 108 or a battery of high-speed projectors.
- the data generated in step S130 may be appropriately reduced again by quantization, but at the cost of data loss.
- a region of interest encoding Rol coding
- the amount of data can be binary coded and further reduced, e.g. by run-length coding and similar approaches known in the art.
- the transmission in step S132 may be over the Internet or an intranet, for example, through a client-server relationship using the TCP-IP protocol, the UDP protocol, or the server-side protocol.
- the transmission S132 can also lead to a local storage of the received frames / frames as a local file.
- step S134 prior to providing the data to the 3D engine 210, the data may be unpacked prior to transmission and buffering the data, buffers being provided to ensure that after reaching an initial desired filling state continuous processing of the data packets is possible, even at different or varying data rates with which the corresponding data packets are received at the receiver.
- step S138 may include displaying using the stereoscopic 3D Pepper's ghost method to generate holograms, as illustrated in step S140, where an autostereoscopic SD monitor or a battery of high-speed projectors is provided, in order to project suitable SD images, for example the images 100a, 100b (see FIG. 2) generated by the 3D engine 210, onto a glass pane comprising a lenticular grid or a suitable 3D structure, so that in a predetermined viewing direction the glass pane for the human eye creates a 3D hologram.
- suitable SD images for example the images 100a, 100b (see FIG. 2) generated by the 3D engine 210
- a user selects a perspective on the receiver side from which the object is to be viewed, as indicated in step S142, and according to step S144 the object is subsequently represented from the selected perspective, based on the received data describing the 3-D model of the object without requiring a return channel to the point where the 3-D image of the object is generated.
- step S142 the object is subsequently represented from the selected perspective, based on the received data describing the 3-D model of the object without requiring a return channel to the point where the 3-D image of the object is generated.
- the recipient side has the option of a user freely selecting a perspective with which he wants to look at the corresponding object Representation required 3D recordings generated from the 3 D model on the receiver side.
- a perspective with which he wants to look at the corresponding object Representation required 3D recordings generated from the 3 D model on the receiver side.
- the corresponding 3D images for the representation on the receiver-side monitor can be generated without the need for a new recording and thus a retransmission of the data or a return channel to the transmitter.
- This embodiment is preferably used in the following applications: ⁇ Presentations at trade fairs or shows with glass panel in human size or larger;
- Communication application in the desk e.g. in the form of a monitor-sized glass panel that can be moved out of the desk, on which a hologram of the face of a conversation partner appears;
- ⁇ In the living area, in which a glass pane with projection hardware of any size is set up, or outside window panes, so that one or more conversation partners standing in the living area can be displayed as a hologram.
- the present embodiments are applied to monitoring as display and transmission of changing contents. For some applications, monitoring, detection, and transmission of changing content is particularly important.
- differences of a static 3 D model within certain limits (thresholds) are created to a recorded live image to detect changes faster and more accurate than in a 2D video image. For example, consider the monitoring of a drilling rig.
- a static 3 D model of the drilling rig is compared several times a second with a 3D image from a viewpoint on the drilling rig, whereby, for example, the drilling rig can be animated via the 3-D engine during runtime.
- Changes in the live 3D model such as entering a human into a shooting area, are compared to the static SD model and can trigger alerts.
- embodiments provide a method for generating data for a two- or three-dimensional representation of at least a portion of an object, comprising: Generating a 3-D model of the part of the object using an SD photograph of the part of the object; and
- the approach for generating data for a two- or three-dimensional representation is advantageous since, unlike in conventional approaches, the elaborate transmission of two-dimensional image data is dispensed with. Rather, starting from the three-dimensional image data, which represent a 3D image of the part of the object, a 3 D model is created, which represents at least the part of the object from which the SD image was obtained.
- This 3D model can be, for example, a grid model or a triangular mesh, as is known, for example, from the field of CAD technology.
- the model obtained in this way can be described by the position of the vertex values in three-dimensional space, for example in the Cartesian coordinate system by the X, Y and Z values of the vertices.
- the color values can be assigned to the corresponding vertices, if necessary also texture information is transmitted.
- the amount of data generated thereby is many times less than the amount of data obtained when transferring a 2D image with a size of 1024 x 768 pixels, so that due to the small amount of data for displaying the object in three-dimensional form and delay-free transmission of the data over a transmission medium is made possible, in particular the problems associated with the conventionally incurred in the prior art large amounts of data are prevented.
- the data thus generated can either be used to generate a three-dimensional representation (eg a 3D I_ivesequenz or a 3D movie) or to generate a two-dimensional representation (eg a 2 D-Livesequenz or a 2D film) on a suitable display device for displaying the object or part of the object.
- a three-dimensional representation eg a 3D I_ivesequenz or a 3D movie
- a two-dimensional representation eg a 2 D-Livesequenz or a 2D film
- the 3D capture includes the portion of the object and a background, the method further comprising extracting the background from the data using the Z value of each vertex, for example, by removing a vertex from the data when the Z value the peak value is outside a predefined range.
- the method may additionally be provided to correct the edge region of the object by filtering out depth distances exceeding a predetermined threshold.
- This procedure is advantageous because it allows the recorded object to be displayed in a simple manner without the likewise recorded background, and thus only the data for the actual object is generated, but not the background data that may not even be required, so that This achieves a further reduction in the amount of data. Furthermore, this procedure makes it possible to represent the object to be displayed three-dimensionally by the generated data on the receiving side in another context, for example in front of another background.
- the 3D model is generated using at least a first 3D image and a second 3D image of the object from respectively different positions, the first and second 3D images overlapping at least partially.
- the different positions may be selected such that a region of the object that is not visible in the first 3D image of the object is visible in the second 3D image of the object.
- it may be provided to generate a first 3D model using the first 3D image and a second 3D model using the second 3D image, and to add the first and second 3D models into a common 3D model connect, providing the data using the common 3D model.
- connecting the first and second 3D models to a common 3D model may include: arranging the first and second SD models such that their overlapping areas are in registration, identifying the vertices from the first 3D model, and from the second 3D model lying within a plane within a predefined distance, and merging the identified vertices into a new vertex in the common 3D model.
- the identifying and summarizing is repeated for a plurality of levels, wherein the number and the spacing of the plurality of levels is selected so that the part of the object is represented by the common 3D model.
- This refinement is advantageous since it ensures that all regions of the part of the object which is to be represented three-dimensionally are described by corresponding data of the 3D model, so that, in particular in a representation, not shown sections of the object can be seen from FIG different perspective and from a different perspective.
- the provision of the data comprises reducing the amount of data without data loss, by determining the spatial distances of the peak values, and correcting the spatial distances as a sequence difference starting from a predetermined starting point. According to embodiments, it may be provided to start coding at a lower point of the object and to continue spiraling to an upper point of the object.
- This procedure is advantageous since it allows a further reduction in the amount of data without any loss of data since, starting from the starting point which is completely encoded, only the difference values to the positions of the adjacent vertices have to be defined, which leads to the further reduction mentioned the amount of data.
- the generation of the 3 D model comprises providing the 3D image of the object or the part of the object by a 3D camera or providing a stereoscopic image of the object or the part of the object.
- the object is a person, an animal, an object or a background.
- This approach is advantageous because the approach has not been subject to any restrictions with regard to the object to be displayed, since from the 3D images of the object in the manner described a significantly reduced data set for describing the same can be obtained in three-dimensional form.
- the object is a background
- this is advantageous since it provides the possibility of providing a desired background and of providing it to a remote position for display in a three-dimensional configuration, in particular the representation of the entire background, for example one Space taken by several cameras in a three-dimensional way, the creation of data to represent the background based on the 3D model, which are transmitted with a reduced amount of data and allow it at the receiving point, the background, such as the room to be generated in such a way that a viewer at the place of reception is able to perceive the background from any position / perspective.
- generating the 3D model and providing the data at a predetermined repetition rate are repeated to generate a plurality of temporally consecutive frames, each containing the provided data and displayable as a 3-D sequence.
- the repetition rate is selected such that up to 30 frames are generated in one second.
- the object is at least the face of a person
- the present method comprises for each frame the following: see facial model of an original face of another person, determining a position of the face of the person in the room in the generation of the 3 D shot, overlaying the 3D model of the person's face with the static face model of the other person, fitting the 3D model of the face the person in those places where there is no movement, to the static face model of the other person, creating a texture from the 3D image of the person's face, which is transparent at the respective locations where there is movement, to create a hole mask texture, and semi-transparent texturing of the hole mask texture to the customized 3D model of the person's face so that a resulting 3D sequence to the human eye shows a moving and animated representation of the original face.
- This approach is advantageous in that it allows a simple way of associating a person resembling a known person in terms of physique and stature with the facial contours of the other person, so that the data provided are based on those described Have a small amount of data, can be prepared at the receiving point so that the local viewer gets the impression that the other person is shown, which is particularly advantageous in the field of entertainment industry and similar areas.
- Embodiments also provide a method of generating a two- or three-dimensional representation of at least a portion of an object, comprising:
- This procedure is advantageous because, due to the described approach, the data generated for the three-dimensional representation have only a small amount of data and are therefore transmitted in a simple manner.
- the data thus received, which display the 3D model, make it possible, by applying a 3D engine, to generate a corresponding 3D image of the part of the object, which then opens conventional manner for three-dimensional representation on a display device, such as a stereoscopic monitor can be used.
- a display device such as a stereoscopic monitor
- the 3D engine generates up to 30 3D images per second, which is advantageous since moving images, ie 3 D films or 3 D sequences, can be generated from the received data.
- the method includes displaying the 3D shot by a display device, e.g. an autostereoscopic 3D monitor or by a battery of high-speed projectors, wherein it may be provided that the display device using the stereoscopic 3D Pepper's Ghost method for generating holograms works.
- the method may comprise projecting the 3D images generated by the 3D engine onto a glass pane comprising a lenticular grid or a suitable 3D structure such that a 3D hologram is provided within a region in front of the glass pane for the human eye arises.
- the transmission of the data may include transmission over the Internet or an intranet, e.g. through a client-server relationship using, for example, the TCP-IP, UDP, or Serverside protocol.
- a local storage of the data may be provided as a file.
- This procedure is advantageous because conventional transmission media are used since, due to the described procedure, there are no special requirements for the provided bandwidth and transmission capacity.
- This approach is a particular advantage of the approach described because a 3D model of the object or part of the object is generated at the receiver side, which is transmitted in the manner described above to the receiver side, with a reduced amount of data. This allows the receiver side, based on a corresponding 3D engine, to recover the original 3D model so that the entire 3D representation of the object is present at the receiver side.
- Such a procedure is not readily possible in the state of the art, as here only the images generated on the transmitter side are present, which are to be combined for the three-dimensional representation on the display.
- Embodiments further provide a computer program with instructions for carrying out the method when the instructions are executed by a computer, whereby the advantages mentioned above are also achieved in a computer implementation of the described approach.
- Embodiments further provide apparatus for generating data for a two- or three-dimensional representation of at least a portion of an object, comprising: an input configured to receive a 3D photograph of the portion of the object; a 3D engine connected to the input and configured to generate a 3D model of the part of the object using the received 3D image and, using the 3D model, to generate data representing the vertices of the part of the object in three-dimensional space and the
- Describe color values associated with vertices and an output connected to the 3D engine and configured to provide the generated data.
- the 3D engine is configured to extract a background in the 3D image from the data using the Z value of each vertex, preferably an edge region of the part of the object by filtering out depth distances exceeding a predetermined threshold. Getting corrected.
- the input is configured to receive at least a first SD image and a second 3D image of the object from respectively different positions, the first and second 3D images at least partially overlapping.
- the 3D engine is configured to generate the 3D model using the first 3D image and the second 3D image of the object.
- the 3D engine generates a first 3D model using the first 3D shot and a second 3D model using the second 3D shot and a common 3D model using the first and second 3D models, wherein the data is generated using the common 3D model.
- the 3D engine effects a reduction in the amount of data without loss of data by determining the spatial distances of the peaks, and coding the spatial distances as a sequence, starting from a given starting point to an end point.
- the device comprises a 3D camera or a stereoscopic recording device for generating the 3D image of the object, wherein the 3D camera or the stereoscopic recording device is connected to the input.
- the 3D engine is configured to generate the 3D model and data at a particular repetition rate to generate a plurality of temporally consecutive frames, each containing the provided data and displayable as a 3D sequence.
- the object is at least a person's face
- the 3D engine is configured to determine, for each frame, a position of the person's face in space in the generation of the 3D shot, the 3D model of the person's face overlaying another person's face with a static facial model of an original face in order to fit the 3D model of the person's face in those areas where there is no movement to the static facial model of the other person in order to obtain a texture from the 3D
- Embodiments provide a system for generating a two- or three-dimensional representation of at least a portion of an object, comprising: means for generating data for the two- or three-dimensional representation of the at least one portion of the object according to embodiments, a transmission medium for transmitting the data is connected to the device for generating the data, and a 3D engine connected to the transmission medium and configured to generate a 2D or 3D photograph of the portion of the object using the transmitted data.
- the system comprises a display device, for example in the form of an autostereoscopic 3D monitor or in the form of a battery of high-speed projectors, which is connected to the 3D engine.
- the display device operates using the stereoscopic 3D Pepper's Ghost method for generating holograms, and is configured to project the 3D images generated by the 3D engine onto a glass panel that includes a lenticular grid or a suitable 3D structure. so that a 3D hologram is created for the human eye within an area in front of the glass pane.
- the 3D engine of the system is configured to receive a selection of a perspective from which to view the object and the object from the selected perspective based on the received data representing the 3D model of the object describe, so that no return channel is required to the point where the 3D image of the object is generated.
- Embodiments thus provide the opportunity for interactive real-time SD graphics.
- the existing problem in the prior art with regard to the enormous amount of data concerning the representation of 3D objects is addressed in accordance with embodiments with the aid of a 3D engine on which, for example, a so-called real-time software for the representation of a computer graphic runs.
- the spatial component of the 3D model is exploited as a deterministic sequence of spatial distances, so that the amount of data is significantly reduced while the quality remains the same.
- the resulting amount of data as a result of spatial distances is orders of magnitude smaller than the amount of data in the prior art for transmitting 3D image data as a result of 2D images.
- the spatial encoding of distance values can be performed faster than the compression of 2D image data, making it possible to perform a live transmission of 3D image data in real time, with a suitable 3D engine at the receiver in three dimensions with interactive viewpoint change or as a stereoscopic movie as multiple 2D images per frame.
- Embodiments have been described in connection with the 3D representation of an object or a part of the object.
- the described approach can also be used in a two-dimensional representation of the object or part of the object, e.g. in that the generated data, which reproduce the SD model, are processed and displayed on the receiver side only as 2D images or as a 2D image sequence. Also in this case, the above-mentioned advantages result over a conventional transmission of 2D images.
- aspects have been described in the context of a device, it will be understood that these aspects also constitute a description of the corresponding method, so that a block or a component of a device is also to be understood as a corresponding method step or as a feature of a method step. Similarly, aspects described in connection with or as a method step also represent a description of a corresponding block or detail or feature of a corresponding device. Depending on particular implementation requirements, embodiments of the invention may be implemented in hardware or in software.
- the implementation may be performed using a digital storage medium, such as a floppy disk, a DVD, a Blu-ray Disc, a CD, a ROM, a PROM, an EPROM, an EEPROM or FLASH memory, a hard disk, or other magnetic disk or optical memory are stored on the electronically readable control signals that can cooperate with a programmable computer system or cooperate such that the respective method is performed. Therefore, the digital storage medium can be computer readable.
- some embodiments according to the invention include a data carrier having electronically readable control signals capable of interacting with a programmable computer system such that one of the methods described herein is performed.
- embodiments of the present invention may be implemented as a computer program product having a program code, wherein the program code is operable to perform one of the methods when the Computer program product runs on a computer.
- the program code can also be stored, for example, on a machine-readable carrier.
- inventions include the computer program for performing any of the methods described herein, wherein the computer program is stored on a machine-readable medium.
- an embodiment of the method according to the invention is thus a computer program which has a program code for performing one of the methods described herein when the computer program runs on a computer.
- a further embodiment of the inventive method is thus a data carrier (or a digital storage medium or a computer-readable medium) on which the computer program is recorded for carrying out one of the methods described herein.
- a further embodiment of the method according to the invention is thus a data stream or a sequence of signals, which represent the computer program for performing one of the methods described herein.
- the data stream or the sequence of signals may be configured, for example, to be transferred via a data communication connection, for example via the Internet.
- Another embodiment includes a processing device, such as a computer or a programmable logic device, that is configured or adapted to perform one of the methods described herein.
- a processing device such as a computer or a programmable logic device, that is configured or adapted to perform one of the methods described herein.
- Another embodiment includes a computer on which the computer program is installed to perform one of the methods described herein.
- a programmable logic device eg, a field programmable gate array, an FPGA
- a field programmable gate array may cooperate with a microprocessor to perform one of the methods described herein.
- the methods are performed by any hardware device. These may be a universal hardware such as a computer processor (CPU) or hardware specific to the process, such as an ASIC.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Processing Or Creating Images (AREA)
- Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
Abstract
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
DE102017203721.0A DE102017203721A1 (de) | 2017-03-07 | 2017-03-07 | Vorrichtung und Verfahren zur Darstellung eines Raumbilds von einem Objekt in einer virtuellen Umgebung |
PCT/EP2018/055518 WO2018162509A2 (fr) | 2017-03-07 | 2018-03-06 | Dispositif et procédé pour représenter un modèle d'un objet dans un environnement virtuel |
Publications (1)
Publication Number | Publication Date |
---|---|
EP3593528A2 true EP3593528A2 (fr) | 2020-01-15 |
Family
ID=63259000
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP18764183.2A Pending EP3593528A2 (fr) | 2017-03-07 | 2018-03-06 | Dispositif et procédé pour représenter un modèle d'un objet dans un environnement virtuel |
Country Status (4)
Country | Link |
---|---|
US (1) | US11652970B2 (fr) |
EP (1) | EP3593528A2 (fr) |
DE (1) | DE102017203721A1 (fr) |
WO (1) | WO2018162509A2 (fr) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20230110832A (ko) | 2017-12-22 | 2023-07-25 | 매직 립, 인코포레이티드 | 가상, 증강 또는 혼합 현실 환경에서 3d 비디오들을생성 및 디스플레이하기 위한 방법들 및 시스템 |
CN114616821B (zh) * | 2019-10-28 | 2023-03-24 | 奇跃公司 | 用于虚拟对象呈现的边缘检测和平滑处理 |
CN111369687B (zh) * | 2020-03-04 | 2021-03-30 | 腾讯科技(深圳)有限公司 | 合成虚拟对象的动作序列的方法及设备 |
CN111464828A (zh) * | 2020-05-14 | 2020-07-28 | 广州酷狗计算机科技有限公司 | 虚拟特效显示方法、装置、终端及存储介质 |
US11538214B2 (en) * | 2020-11-09 | 2022-12-27 | Meta Platforms Technologies, Llc | Systems and methods for displaying stereoscopic rendered image data captured from multiple perspectives |
US20240096033A1 (en) * | 2021-10-11 | 2024-03-21 | Meta Platforms Technologies, Llc | Technology for creating, replicating and/or controlling avatars in extended reality |
Family Cites Families (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH09153146A (ja) * | 1995-09-28 | 1997-06-10 | Toshiba Corp | 仮想空間表示方法 |
US6873340B2 (en) * | 1997-05-15 | 2005-03-29 | Visimatix, Inc. | Method and apparatus for an automated reference indicator system for photographic and video images |
JP4772544B2 (ja) * | 2005-04-27 | 2011-09-14 | 富士フイルム株式会社 | 撮像装置、撮像方法、及びプログラム |
US8094928B2 (en) * | 2005-11-14 | 2012-01-10 | Microsoft Corporation | Stereo video for gaming |
US20070210985A1 (en) * | 2006-03-13 | 2007-09-13 | Royer George R | Three dimensional imaging system |
DE102007028654A1 (de) * | 2007-06-19 | 2009-01-15 | Institut für Arbeitsphysiologie an der Universität Dortmund | Vorrichtung und Verfahren zur ergonomischen Darstellung der räumlichen Tiefe von Objekten in virtuellen Umgebungen |
US8743176B2 (en) * | 2009-05-20 | 2014-06-03 | Advanced Scientific Concepts, Inc. | 3-dimensional hybrid camera and production system |
US8368696B2 (en) * | 2009-06-19 | 2013-02-05 | Sharp Laboratories Of America, Inc. | Temporal parallax induced display |
US9013559B2 (en) * | 2010-02-02 | 2015-04-21 | Konica Minolta Holdings, Inc. | System, method and program for capturing images from a virtual viewpoint |
US20130278631A1 (en) * | 2010-02-28 | 2013-10-24 | Osterhout Group, Inc. | 3d positioning of augmented reality information |
US8438590B2 (en) * | 2010-09-22 | 2013-05-07 | General Instrument Corporation | System and method for measuring audience reaction to media content |
US20120130717A1 (en) * | 2010-11-19 | 2012-05-24 | Microsoft Corporation | Real-time Animation for an Expressive Avatar |
DE102011008886A1 (de) * | 2011-01-19 | 2012-07-19 | Sterrix Technologies UG (haftungsbeschränkt) | Verfahren und Vorrichtung zur Stereobasis-Erweiterung von stereoskopischen Bildern und Bildfolgen |
US9292959B2 (en) * | 2012-05-16 | 2016-03-22 | Digizig Media Inc. | Multi-dimensional stacking with self-correction |
US10008027B1 (en) * | 2014-10-20 | 2018-06-26 | Henry Harlyn Baker | Techniques for determining a three-dimensional representation of a surface of an object from a set of images |
US11956414B2 (en) * | 2015-03-17 | 2024-04-09 | Raytrx, Llc | Wearable image manipulation and control system with correction for vision defects and augmentation of vision and sensing |
US10555021B2 (en) * | 2015-08-31 | 2020-02-04 | Orcam Technologies Ltd. | Systems and methods for selecting content based on a user's behavior |
CN106999024B (zh) * | 2015-09-01 | 2018-09-28 | 奥林巴斯株式会社 | 摄像*** |
GB2544971B (en) * | 2015-11-27 | 2017-12-27 | Holition Ltd | Locating and tracking fingernails in images |
CN107643603B (zh) * | 2016-07-21 | 2021-04-09 | 天马微电子股份有限公司 | 立体显示装置及控制方法 |
US10735691B2 (en) * | 2016-11-08 | 2020-08-04 | Rockwell Automation Technologies, Inc. | Virtual reality and augmented reality for industrial automation |
EP3343432B1 (fr) * | 2016-12-29 | 2024-03-20 | Elektrobit Automotive GmbH | Génération des images à training pour systèmes d'identifications d'objects basé sur l'apprendre automatisé |
-
2017
- 2017-03-07 DE DE102017203721.0A patent/DE102017203721A1/de active Pending
-
2018
- 2018-03-06 WO PCT/EP2018/055518 patent/WO2018162509A2/fr unknown
- 2018-03-06 EP EP18764183.2A patent/EP3593528A2/fr active Pending
-
2019
- 2019-09-05 US US16/561,341 patent/US11652970B2/en active Active
Also Published As
Publication number | Publication date |
---|---|
US11652970B2 (en) | 2023-05-16 |
DE102017203721A1 (de) | 2018-09-13 |
WO2018162509A3 (fr) | 2020-01-02 |
WO2018162509A2 (fr) | 2018-09-13 |
US20190394443A1 (en) | 2019-12-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP3347876B1 (fr) | Dispositif et procédé pour générer un modèle d'un objet au moyen de données-image de superposition dans un environnement virtuel | |
EP3304496B1 (fr) | Procédé et dispositif pour générer des données pour une représentation à deux ou trois dimensions d'au moins une partie d'un objet et pour générer la représentation à deux ou trois dimensions de la partie de l'objet | |
EP3593528A2 (fr) | Dispositif et procédé pour représenter un modèle d'un objet dans un environnement virtuel | |
DE69932619T2 (de) | Verfahren und system zum aufnehmen und repräsentieren von dreidimensionaler geometrie, farbe und schatten von animierten objekten | |
DE112018001822T5 (de) | Skelettbasierte Effekte und Hintergrundaustausch | |
DE112018000311T5 (de) | Stereoskopisches Rendering unter Verwendung von Raymarching und ein Broadcaster für eine virtuelle Ansicht für solches Rendering | |
DE112016004640T5 (de) | Filmische bearbeitung für virtuelle realität und erweiterte realität | |
US9582918B2 (en) | Techniques for producing creative stereo parameters for stereoscopic computer animation | |
CN112446939A (zh) | 三维模型动态渲染方法、装置、电子设备及存储介质 | |
DE112015004871T5 (de) | Verfahren zur automatisierten Echtzeit-Konversion von 2D-RGB-Bildern und -Video in stereoskopische Rot-Cyan-3D-Anaglyphen | |
US20220114784A1 (en) | Device and method for generating a model of an object with superposition image data in a virtual environment | |
WO2017085325A1 (fr) | Dispositif et procédé pour superposer au moins une partie d'un objet à une surface virtuelle | |
EP3057316B1 (fr) | Génération d'imagerie tridimensionnelle en complément du contenu existant | |
WO2009118156A2 (fr) | Procédé de production d'une image 3d d'une scène à partir d'une image 2d de la scène | |
WO2024055211A1 (fr) | Procédé et système de reconstruction vidéo en trois dimensions sur la base d'une combinaison nerf de couches à vues multiples | |
EP2478705A1 (fr) | Procédé et dispositif de production d'images partielles et/ou d'un modèle d'image stéréoscopique à partir d'une vue 2d pour une reproduction stéréoscopique | |
US20210327121A1 (en) | Display based mixed-reality device | |
DE102015017128A1 (de) | Verfahren und Vorrichtung zum Erzeugen von Daten für eine zwei- oder dreidimensionale Darstellung zumindest eines Teils eines Objekts und zum Erzeugen der zwei- oder dreidimensionalen Darstellung zumindest des Teils des Objekts | |
JP2009500878A (ja) | 奥行錯覚デジタル撮像 | |
DE112021003549T5 (de) | Informationsverarbeitungsvorrichtung, informationsverarbeitungsverfahren und programm | |
Caviedes et al. | Real time 2D to 3D conversion: Technical and visual quality requirements | |
KR20210090180A (ko) | 화상 처리 디바이스, 화상 처리 방법, 프로그램, 및 표시 디바이스 | |
US7212662B2 (en) | Method for the compressed transmission of image data for 3-dimensional representation of scenes and objects | |
DE112022004369T5 (de) | Erfassung und wiedergabe von umgebungen | |
DE112020001322T5 (de) | Eine szene darstellendes bildsignal |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
17P | Request for examination filed |
Effective date: 20190905 |
|
AK | Designated contracting states |
Kind code of ref document: A2 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
AX | Request for extension of the european patent |
Extension state: BA ME |
|
R17D | Deferred search report published (corrected) |
Effective date: 20200102 |
|
RIC1 | Information provided on ipc code assigned before grant |
Ipc: H04N 13/275 20180101ALI20200824BHEP Ipc: H04N 7/18 20060101ALI20200824BHEP Ipc: G06T 7/50 20170101AFI20200824BHEP Ipc: H04N 13/239 20180101ALI20200824BHEP |
|
DAV | Request for validation of the european patent (deleted) | ||
DAX | Request for extension of the european patent (deleted) | ||
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
17Q | First examination report despatched |
Effective date: 20210420 |