Graphical Diagram for Video Object Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video navigation methods, such as fast-forwarding or using prespecified scenes, are inefficient for locating specific portions of a video, especially when audio is involved, and lack sufficient granularity to select desired video segments without knowing their chronological location.
Innovation Solution
A graphical diagram is generated with elements representing objects from a video, each element's size and location corresponding to the number and relationships of video portions where the object appears, allowing users to select portions by choosing points or areas within the diagram that overlap with desired objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional navigation methods (fast-forwarding or prespecified scenes) are used, then the system is simple to operate, but the efficiency of locating specific video portions is poor
Solution Approach 1:
The patent transforms the traditional linear chronological navigation into a two-dimensional spatial representation. The timeline is extended vertically to show multiple objects simultaneously, allowing users to locate video portions by spatial positioning rather than sequential searching. This dimensional transformation enables parallel visualization of multiple objects across different time points, dramatically improving location efficiency without requiring complex multi-step operations.
Solution Approach 2:
The patent segments the video content by individual objects, creating separate graphical elements for each object that appears in the video. Each object is represented as an independent element with its own temporal span, allowing users to selectively interact with specific objects. This segmentation enables precise targeting of desired video portions without navigating through entire scenes or relying on prespecified breakpoints.
2Measurement precision
If prespecified scenes are used for navigation, then the device complexity is low, but the granularity for selecting video segments is insufficient
Solution Approach 1:
The patent divides the video into object-based segments rather than using fixed prespecified scenes. Each object generates its own graphical element representing its temporal presence, creating fine-grained segments that naturally align with meaningful content units. This segmentation provides precise control over video portion selection, allowing users to target exactly the segments containing their objects of interest without being constrained by coarse scene boundaries.
Solution Approach 2:
The patent creates a dynamic navigation system where the granularity of video segments adapts to the content itself. Rather than using fixed prespecified breakpoints, the system automatically generates segment boundaries based on object appearance and disappearance times. This dynamic segmentation provides variable precision - finer where objects appear/disappear and coarser where objects persist - matching the user's selection needs without manual configuration.
3Ease of operation
If users must know chronological locations to select video portions, then the navigation system is simple, but the ease of operation for locating objects of interest is reduced
Solution Approach 1:
The patent adds a spatial dimension to the temporal information by displaying multiple objects vertically across the timeline. This transformation allows users to locate objects of interest through spatial recognition and selection rather than requiring knowledge of chronological positions. The chronological context is preserved through the horizontal timeline axis, while the vertical dimension enables direct object-based navigation, eliminating the need to remember or estimate time positions.
Solution Approach 2:
The patent creates graphical element copies that represent each object's presence in the video. These visual copies serve as proxies for the actual video content, allowing users to interact with simplified representations rather than the full video timeline. Each graphical element encapsulates the temporal and spatial information of its corresponding object, enabling selection based on object identity rather than chronological location knowledge.
Data Source
AI summary
A video is divided into portions and has objects. Each object appears in one or more portions. A graphical diagram is generated and displayed that has graphical elements corresponding to the objects. Each graphical element has a size corresponding to a number and/or size of the portions in which the object to which the graphical element corresponds appears. The graphical elements have locations within the graphical diagram in correspondence with relationships of appearance of the objects within the portions. Selection of a point or area within the graphical diagram is received. Which of the graphical elements include the point or area are determined as one or more selected graphical elements corresponding to one or more selected objects. Which of the portions include all the selected objects are determined and output as one or more selected portions.