Contextual Visual Search Using Depth-Based Object Highlighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing scanning systems fail to effectively distinguish and highlight specific objects within 3D representations, making it difficult to efficiently search for desired items in a physical environment.
Innovation Solution
Utilizing depth-based data to determine a searched-for object and applying diminished reality effects to visually distinguish it from other objects, such as by removing, blurring, or making objects translucent, thereby enhancing the search process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If all objects are displayed with equal visual prominence in 3D representations, then complete environmental information is preserved, but object search efficiency deteriorates due to visual clutter
Solution Approach 1:
The patent applies different visual properties to different objects based on their relevance to the search query. The highlighted object receives enhanced visual treatment (brighter colors, increased saturation) while other objects maintain normal or reduced visual properties, creating local quality differences that guide user attention without removing contextual information
Solution Approach 2:
The system changes the color properties of objects to encode search relevance information. The target object is highlighted through color modifications (increased brightness, saturation, or contrasting hues) while other objects may have their colors muted or desaturated, allowing users to quickly identify the search target while preserving spatial and contextual awareness
2Loss of time
If visual effects are applied to highlight the searched-for object, then object identification speed is improved, but system complexity increases
Solution Approach 1:
The system uses relatively simple color space transformations (adjusting brightness, saturation, and hue parameters) to highlight objects rather than complex visual effects. This approach achieves rapid object identification through computationally efficient color modifications that can be applied in real-time without requiring heavy processing resources
Solution Approach 2:
The visual processing is segmented into distinct stages: object detection and identification, relevance determination, and selective visual property modification. This segmentation allows the system to apply visual effects only to specific objects that match search criteria, reducing overall processing complexity compared to applying effects to the entire scene
Data Source
AI summary
Various implementations disclosed herein include devices, systems, and methods that provide a view of a physical environment with a visual effect based on determining whether an object is a searched-for object. For example, an example process may include determining an objective corresponding to identifying a searched-for object in a physical environment including one or more objects. The process may further include obtaining depth-based data based on sensor data captured by one or more sensors in the physical environment. The process may further include determining, based on the determined objective and the depth-based data, whether a first object in the physical environment is the searched-for object. The process may further include, in accordance with determining whether the first object is the searched-for object, providing a view of the physical environment with a visual effect.