Prior art — US Patent 10769446

Here is a detailed analysis of the prior art cited in U.S. Patent No. 10,769,446, focusing on the most relevant references and their potential impact on the patent's claims.

Analysis of Prior Art for U.S. Patent No. 10,769,446

The following analysis details the prior art references cited during the prosecution of U.S. Patent No. 10,769,446. Each reference is assessed for its relevance to the patent's independent claims, particularly claims 1, 9, and 15, which describe a method, system, and non-transitory computer-readable medium for combining video content with augmentations based on user interaction with defined bounding boxes.

Most Relevant Prior Art

1. U.S. Patent No. 9,495,788 B2 ("the '788 patent")

Full Citation: US 9,495,788 B2, "System and method for augmenting a broadcast video with secondary content related to a user-selected object"
Assignee: Rovi Guides, Inc.
Filed: March 13, 2013
Issued: November 15, 2016
Brief Description: The '788 patent describes a system that identifies objects in a video stream, associates them with supplementary information, and allows a user to select an object on-screen to retrieve and display that information. It explicitly mentions identifying an object's location, receiving a user selection corresponding to that location, and presenting related content.
Potential Anticipation Analysis:
- Claim 1 (Method): This reference appears to disclose most elements of claim 1. It teaches displaying a video, identifying objects (which implies a form of bounding box or coordinate system), detecting a user's selection of an object, and displaying augmented information related to that object. The key difference may lie in the specific method of identifying the selected box using RGBA values, which is a specific implementation detail in the '446 patent. However, the broader process is substantially described.
- Claim 9 (System): The system described in the '788 patent, involving a user device that receives video, processes user input, and displays augmented content, aligns closely with the system defined in claim 9.
- Claim 15 (Server-Side Method): The '788 patent discloses a system where a server can provide the video and associated metadata (e.g., object locations and related content) to a client device. It describes the client sending a selection back and the system providing the relevant secondary content. This mirrors the server-client interaction described in claim 15.
Conclusion: The '788 patent is highly relevant and could be argued to anticipate the core concept of claims 1, 9, and 15. The novelty of the '446 patent may rest on the specific technical implementation, such as the use of RGBA values for identifying transparent bounding boxes.

2. U.S. Patent No. 9,986,227 B2 ("the '227 patent")

Full Citation: US 9,986,227 B2, "Augmented reality system including a broadcast production system for a live event"
Assignee: Replay Technologies Ltd.
Filed: October 26, 2015
Issued: May 29, 2018
Brief Description: This patent details a system for creating augmented reality views of live events, particularly sports. It involves capturing video from multiple angles, generating a 3D model of the event, and overlaying graphical information. The system allows users to interact with the broadcast, for example, by selecting a player to see their statistics.
Potential Anticipation Analysis:
- Claim 1 (Method): The '227 patent describes presenting video of an event and allowing user selection of objects (players) to display augmentations (statistics). This covers the fundamental process of claim 1. While it focuses on a 3D reconstructed environment, the core interaction model is similar. The method for linking a user's screen tap to a specific player is not detailed as using RGBA values but is functionally equivalent.
- Claim 9 (System): The client device in the '227 patent's system is designed to receive video, detect user input on displayed objects, and render augmented reality content, which is analogous to the system described in claim 9.
- Claim 15 (Server-Side Method): The '227 patent's architecture relies on a server-side component (the "broadcast production system") to process video feeds, identify objects, and manage the data for augmentations, which is then sent to client devices. This aligns with the server-side process of claim 15.
Conclusion: The '227 patent is also highly relevant. It describes a sophisticated system for real-time sports video augmentation with user interaction. The patentability of the '446 invention likely hinges on the specific, more efficient method of using invisible bounding boxes and RGBA values for object identification, as opposed to the more complex 3D modeling and object recognition described in the '227 patent.

3. U.S. Patent Application Publication No. 2013/0212530 A1 ("Mistry et al.")

Full Citation: US 2013/0212530 A1, "System and Method for Presenting Supplemental Information for Television Content"
Applicant: Pranav Mistry et al.
Filed: February 13, 2012
Published: August 15, 2013
Brief Description: This application discloses a system that allows a user to interact with video content displayed on a television using a second-screen device (e.g., a smartphone). The user can point their device at an object or person on the TV screen, and the system identifies the object and displays related information on the second screen.
Potential Anticipation Analysis:
- Claim 1 (Method): Mistry et al. describe identifying objects within video frames, allowing a user to select an object, and presenting supplemental content. The core interactive loop is present. The mechanism for selection involves image recognition from the second device's camera rather than a direct screen tap on a bounding box, which is a key distinction.
- Claim 9 (System): The system involves a client device for interaction and display, but it operates as a second screen, which differs from the integrated display system described in claim 9 of the '446 patent.
- Claim 15 (Server-Side Method): A server component is described for processing the video, identifying objects, and providing the supplemental information. This has a strong functional overlap with claim 15.
Conclusion: Mistry et al. is relevant as it describes the concept of selecting objects in a video to get more information. However, its reliance on a two-screen system and a different selection mechanism (camera-based pointing) differentiates it from the '446 patent's single-screen, direct-touch, bounding-box-based method. It weakens the non-obviousness of the general concept but may not directly anticipate the specific claims.

Less Relevant Prior Art

The remaining references cited against the patent, while related to the general fields of video analysis, augmented reality, and data visualization, are less directly relevant to the specific combination of claim elements.

U.S. Patent No. 9,491,424 B2 (KudMost et al.): Focuses on synchronizing video content with external data sources but does not detail the specific user interaction mechanism of selecting objects within the video via bounding boxes to trigger augmentations.
U.S. Patent No. 8,935,720 B2 (Ellis et al.): Relates to providing interactive television advertising and supplementary content, but the method of interaction and object identification is not as specific as that claimed in the '446 patent.
U.S. Patent No. 9,697,643 B2 (Hsu et al.): Describes generating augmented reality views by inserting virtual objects into a real-world video feed, which is a different problem from overlaying informational graphics onto existing objects within a video.
U.S. Patent Application Publication No. 2013/0014072 A1 (Gilley et al.): Pertains to a system for presenting an "enhancement layer" over video content, which can be user-customized. While it involves overlaying information, the specific method of identifying selectable objects using invisible, color-coded bounding boxes is not disclosed.

Overall Assessment

The core invention of US 10,769,446 resides in a specific, efficient method for enabling user interaction with objects in a video stream. By pre-defining transparent bounding boxes, each with a unique identifier (encoded as an RGBA value), the system can quickly determine which object a user has selected without complex image processing on the client device at the time of interaction. The client simply has to read the RGBA value at the point of user input and send that identifier to a renderer or server.

The prior art, particularly US 9,495,788 B2 and US 9,986,227 B2, clearly establishes the broader concept of selecting objects in a video (especially a sports broadcast) to view related data. The patentability of the '446 patent's claims, therefore, likely depends on the novelty and non-obviousness of using this specific RGBA-based bounding box identification technique as the enabling mechanism for this interaction. An argument could be made that this is a specific, inventive implementation that offers advantages in speed and processing efficiency over the more general object recognition or 3D modeling systems described in the prior art.