SELECTED REPLACEMENT OF DIGITAL IMAGERY PORTIONS USING AUGMENTED REALITY
1. A process executable by a computing device to automatically replace static imagery in live video content, the process comprising:
- receiving live video content from a camera that captures an event occurring in a physical space;
initially identifying at least one location in the received live video content that represents static imagery for image replacement;
receiving subsequent live video content from the camera that captures the event occurring in the physical space;
identifying dynamic imagery in the subsequent live video occurring within the at least one replacement location;
adapting the subsequent live video to replace the static imagery at the least one replacement location with a replacement image while retaining the dynamic imagery in the subsequent live video as an overlay of the static image; and
providing the adapted subsequent live video comprising the replacement image as an output.
Systems, devices and automated processes augment television broadcasts with digital content that automatically replaces static portions of the captured imagery. In a video broadcast of a live event, for example, signs or other advertisements that are physically present in the event venue can be automatically replaced in television broadcasts of the event by virtual imagery of other signs or objects, as desired.
- 1. A process executable by a computing device to automatically replace static imagery in live video content, the process comprising:
receiving live video content from a camera that captures an event occurring in a physical space; initially identifying at least one location in the received live video content that represents static imagery for image replacement; receiving subsequent live video content from the camera that captures the event occurring in the physical space; identifying dynamic imagery in the subsequent live video occurring within the at least one replacement location; adapting the subsequent live video to replace the static imagery at the least one replacement location with a replacement image while retaining the dynamic imagery in the subsequent live video as an overlay of the static image; and providing the adapted subsequent live video comprising the replacement image as an output.
- View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11)
- 12. A computer system comprising:
an interface to receive live video content from a camera; and a processor configured to automatically replace static imagery in the live video content by performing a process that comprises; receiving live video content from a camera that captures an event occurring in a physical space; providing the adapted subsequent live video as an output.
- View Dependent Claims (13, 14, 15, 16, 17, 18, 19, 20)
This application claims the benefit of U.S. Provisional Application No. 62/652,631, filed on Apr. 4, 2018.
The following discussion generally relates to digital manipulation of video content, particularly broadcasts of live television programs. More particularly, the following discussion relates to the automated digital replacement of certain static portions of imagery that can be present in broadcast television content.
Digital technology continues to revolutionize television viewing. In addition to the wealth of content that is now available, emerging technologies such as time and place shifting allow viewers to experience increasing amounts of media content at the times and places that are most convenient for them. As viewers gain more control over the content that they view, however, it can be an increasing challenge to maintain the viewer'"'"'s attention.
In recent years, video producers and broadcasters have looked to digital enhancement of live video content (e.g., broadcasts of sporting events) to make the viewer experience more informative and engaging. Most sports broadcasts, for example, incorporate graphical overlays that show the score, time remaining and/or other information to the viewer. Additional enhancements have included emphasizing the puck in broadcasts of hockey matches, highlighting of first down or line of scrimmage positions in broadcasts of football games, and many others. These limited features have been well-received by the viewers, but there is nevertheless a substantial need to make programming more dynamic, and therefore more interesting to the viewer.
It is therefore desirable to create systems and processes to automatically recognize and replace static images in live-captured digital video content. Other desirable features and characteristics will become apparent from the subsequent detailed description and the appended claims, taken in conjunction with the accompanying drawings and this background section.
Various embodiments provide systems, devices and automated processes to augment television broadcasts with digital content that automatically replaces static portions of the captured imagery. Such static portions could include billboards, placards, signs, advertisements, scoreboards and/or other features that are physically present in the stadium or other environment being filmed, and that are therefore viewable in live video content captured during the broadcast event. In a video broadcast of a sporting event, for example, signs or other advertisements that are physically present in the stadium can be temporarily (or permanently) replaced in television broadcasts of the event by virtual imagery of other signs or objects, as desired. Using techniques of augmented reality, image replacement can allow a venue, team, league, broadcaster, content aggregator and/or other party to increase advertising revenue without increasing distractions to the viewer. Automated ad replacement can occur through real-time processing of live video imagery so that there is little or no detraction from the viewer'"'"'s experience. Indeed, by replacing static signs/images with updated imagery throughout the broadcast, the viewer experience can be made even more interesting and engaging.
Some embodiments described below provide processes that are executable by computing devices to automatically replace static imagery in live video content. In one example, a process suitably comprises: receiving live video content from a camera that captures an event occurring in a physical space; initially identifying at least one location in the received live video content that represents static imagery for image replacement; receiving subsequent live video content from the camera that captures the event occurring in the physical space; identifying dynamic imagery in the subsequent live video occurring within the at least one replacement location; adapting the subsequent live video to replace the static imagery at the least one replacement location with a replacement image while retaining the dynamic imagery in the subsequent live video as an overlay of the static image; and providing the adapted subsequent live video as an output.
Other embodiments provide computer systems that comprise an interface to a camera to receive live video content and a processor that performs real-time replacement of static imagery in the live video content. The image replacement may be performed according to any of the various techniques described herein. In one example, the image replacement process performed by the processor of the computer system suitably comprises: receiving live video content from a camera that captures an event occurring in a physical space; initially identifying at least one location in the received live video content that represents static imagery for image replacement; receiving subsequent live video content from the camera that captures the event occurring in the physical space; identifying dynamic imagery in the subsequent live video occurring within the at least one replacement location; adapting the subsequent live video to replace the static imagery at the least one replacement location with a replacement image while retaining the dynamic imagery in the subsequent live video as an overlay of the static image; and providing the adapted subsequent live video as an output.
Further embodiments augment the systems and processes by performing addition analysis and/or manipulation of the captured imagery to match distortion of the imagery caused by moving cameras, to adjust the luminance of the replaced imagery, and/or to otherwise enhance the digital media that is provided to the viewer. These and other examples, aspects, embodiments and other features are described in more detail below.
Example embodiments will hereinafter be described in conjunction with the following drawing figures, wherein like numerals denote like elements, and:
The following detailed description of the invention is intended to provide various examples, but it is not intended to limit the invention or the application and uses of the invention. Furthermore, there is no intention to be bound by any theory presented in the preceding background or the following detailed description.
According to various embodiments, static portions of a video stream that represent signs, advertisements or the like can be automatically replaced with other images. This concept can be used, for example, to replace images of billboards or other advertisements that are physically located within the physical space where a live sporting or other event is occurring, thereby allowing for dynamic updating of the imagery throughout the event. Dynamic replacement may also enable new revenue opportunities for event venues, teams, video producers, broadcasters and/or others as desired.
An additional benefit available in certain implementations is the replacement of local ads with ads that are more relevant to a wider (or at least different) television viewing audience. If a stadium has a prominent sign that advertises a business with only a local market reach (e.g., a hospital or a local restaurant), that particular sign may be relevant to people physically present in the stadium, but not to the wider audience that is viewing the television broadcast. That is, signs that are intended for viewing by people who are physically present at the actual premises of the event may not be relevant to some or all of the television audience viewing a live broadcast of the event. The imagery that represents the sign can therefore be replaced in the television broadcast, thereby leading to more relevant ads, more efficient use of advertising space and potentially increased advertising revenues for the stadium, team, broadcaster or the like.
To that end, various embodiments use digital computing hardware 105 to identify and/or replace certain portions of video imagery corresponding to signs, billboards or other static images within the captured video feed.
In various embodiments, computer system 105 analyzes one or more frames of the captured feed to recognize portions of the image 100 that remain static (e.g., unchanging) throughout the event. In the example illustrated in
Even though space 111 in image 100 is physically painted or otherwise fashioned to represent a particular good or service (e.g., “Brand1” in the illustrated example), the pixels representing space 111 in image 100 can be digitally replaced with different pixels to create a different image 200 (
Camera 102 shown in
Video imagery captured by one or more cameras 102 is provided to a video processing computer system 105. Computer system 105 typically includes one or more digital processors, along with associated memory and input/output interface hardware as would be expected in a conventional video processing computer. System 105 typically executes an automated process 106 to perform the various functions described herein; process 106 may be implemented using computer-executable instructions that are stored in memory for execution by one or more processors, as desired. Computer 105 may be physically located at the premises 101 where the event is occurring; equivalently, computer 105 may be remotely located to operate on a stream of compressed or uncompressed video content that is delivered by a wired or wireless data connection to the camera apparatus, as desired. Although the discussion herein primarily focuses upon live captured video that is processed in real time (or as close to real time as possible, given some limitations that are inherent in image capture, data communications, storage and processing delays, etc.), other embodiments could perform similar functions on previously-recorded digital imagery, as desired.
Processing performed by computer system 105 may include image recognition, image replacement and/or image enhancement, as desired.
With reference to
Static imagery can be recognized in the received video content in any manner (process 304). In various embodiments, pixel values of the received video stream are compared to identify pixels that do not vary from frame to frame. This technique is particularly useful if the video stream is received from a fixed camera 102, or at least a camera 102 that remains relatively motionless during the frame comparison time so that portions of images can be readily compared to each other. By recognizing pixel patterns that remain relatively unchanging over time, occurrences of similar patterns in future images can be automatically recognized as potentially-replaceable static content. Pixel values that are compared could include pixel luminosity in one or more color planes, or any combination of the pixel values from multiple planes (e.g., a sum or average of the pixel'"'"'s values from multiple planes). A further embodiment could expand this concept by averaging pixel values over several frames, and then comparing subsequent pixel values to the previously-determined average, thereby further identifying deviation from an average value.
Further embodiments could automatically average pixel values over time to identify static images that may be occasionally obscured by dynamic action. The floor image 111 in
Typically, computer system 105 will present the recognized areas 111-114 in the received imagery for review by a human operator using an appropriate editing interface. The interface will allow the operator to approve image replacement in the areas, and may also provide a selector to choose desired replacement imagery. Various embodiments could alternately or additionally allow the operator to manually select regions for replacement, if desired, rather than using the automated techniques described herein.
Static areas 111-114 of the captured video imagery are defined in any manner, and stored for subsequent processing (function 306). In various embodiments, the static areas are defined according to pixels in the received imagery; these definitions can be stored for subsequent replacement. Other embodiments may define the replacement portions in terms of the camera position at the time of capture: that is, in addition to storing the location of the relevant portion of the image, the camera location, camera orientation and/or zoom value is stored. If subsequently imagery is captured at the same location and orientation but at a different zoom value, for example, it may be desirable to scale the size of the replaced region to recognize that bigger or smaller position on the screen that is occupied by the static imagery.
Camera orientation may be defined in any manner. In some images, conventional two-dimensional (e.g., “X” and “Y”) coordinates can represent pixels (or ranges of pixels) that can be replaced. In other images, it may be helpful to define the locations in spherical or pseudo-spherical coordinates, as appropriate. In the example of
Static locations 111-114 in the video stream may be identified according to any temporal scheme. In various embodiments, it may be desirable to recognize such locations prior to the broadcast to permit image replacement even at the outset of the broadcast. In embodiments that rely upon automatic detection of static imagery, the recognition process may occur at any regular or irregular interval, in response to user instructions, and/or on an ongoing basis, as desired. Ongoing recognition of static imagery may be computationally intense, but will typically provide the best blending of replacement imagery across a wider array of captured images.
Static images can be replaced in any manner. Various embodiments are able to recognize dynamic imagery that overlies the static images, as appropriate (function 308). This allows for replacement of only those pixels that correspond to the underlying image, with the dynamic content appearing “on top of” the replacement content. As noted above, dynamic content may be recognized by pixel values that deviate from the expected static value, or from an average of the previous values as desired.
To that end, various embodiments will check on a regular (or irregular) basis to confirm that the spaces lying within the previously-recognized static regions of the imagery remain as expected. If the pixel values deviate from frame-to-frame, then the deviated value may be assumed to by dynamic content that overlies the static image.
Typically, replacement images are sized and located to fill the spaces identified in the previously-stored coordinates of the static portions (function 310). One way to accomplish this is by considering pixel values of replacement images: if a static region of XX pixels by YY pixels is recognized, for example, a replacement image can be formatted to be the same size, thereby permitting pixel-level replacement of the static imagery with the replacement imagery.
When a replacement image is desired, the appropriate imagery is scaled or otherwise formatted to fit within the recognized static space of the captured video image (function 312). In various embodiments, the pixels previously-recognized to be presenting dynamic content can remain unchanged in the processed video stream so that the dynamic content continues to be present it the augmented image. That is, only pixels that have the expected value of the static image are replaced; pixels having unexpected values are retained, thereby allowing dynamic content to overlay the replaced imagery, as desired.
Various embodiments could perform additional modification to the replaced image, as desired. Images could be adjusted, for example, to adapt to image distortion due to camera movement or the like, especially within three-dimensional views of the event. If a camera moves through a stadium, for example, captured imagery can be distorted so that the captured imagery may no longer lie within the previously-identified region. Similarly, if a sign or other static image moves relative to the camera during the course of the event, the amorphic shape of the static image may change. If a rectangular object that was previously in the direct view of the camera is moved to a more peripheral position within the camera image, for example, the rectangular object will subsequently appear as a parallelogram with non-parallel sides due to changes in perspective. Various embodiments are therefore able to recognize changes in two- or three-dimensional amorphic imagery due to relative movement of the camera and the object, and to adapt the replacement image accordingly. Changes in amorphic shapes can be recognized through recognition of static image pixel values outside of the expected region (e.g., “image bleed”). Alternatively, changes in region shape can be predicted based upon changes in camera position and/or movements, as desired.
Still further embodiments adapt the replaced imagery to match the luminosity of the captured image. Ideally, the image replacement will be as transparent to the viewer as possible, so it can be beneficial to adapt the image brightness to conform to the surrounding imagery. To provide an example, a sign located in a sports stadium may be relatively bright during a day game when the sign is physically illuminated by bright sunlight. As the sun sets or the sign moves into shade, however, the sign will be less bright in the captured video. It will therefore be beneficial to adapt replacement imagery in a similar manner so that the replaced imagery is not too bright or too dim relative to the surrounding imagery. Various embodiments could monitor changes in luminance for captured pixels lying within the replacement region and adapt the luminance of captured imagery accordingly. Other embodiments could simply retain the luminance values of the captured imagery, but replace the chrominance values of the replaced pixels with those of the replacement image. Still other embodiments could take a hybrid approach, with pixel luminance values being retained but scaled based upon differences between captured and replacement imagery. If a dark-colored billboard is replaced with a light-colored image, for example, it may be desirable to reduce the brightness of the replacement image so the replacement image does not excessively stand out from the surrounding imagery.
The augmented video stream is output in any manner (function 314). In various embodiments, the augmented video stream is supplied for use in a television broadcast that represents the live event occurring in the physical space 101.
As described herein, then, images of real-world advertisements or objects that are physically present in a stadium, theater or other event space can be digitally and automatically replaced with other images, thereby leading to a more dynamic and relevant user experience. These basic concepts may be modified in any number of ways and implemented across a wide array of alternate but equivalent embodiments.
The term “exemplary” is used herein to represent one example, instance or illustration that may have any number of alternates. Any implementation described herein as “exemplary” should not necessarily be construed as preferred or advantageous over other implementations. While several exemplary embodiments have been presented in the foregoing detailed description, it should be appreciated that a vast number of alternate but equivalent variations exist, and the examples presented herein are not intended to limit the scope, applicability, or configuration of the invention in any way. To the contrary, various changes may be made in the function and arrangement of the various features described herein without departing from the scope of the claims and their legal equivalents.