Image-Based Object Search Using Removed Scene Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search systems rely on user navigation and mental visualization to find objects that fit a physical environment, often resulting in user dissatisfaction and inefficient resource use due to mismatches between purchased objects and their surroundings.
Innovation Solution
A computing device processes a digital image of a physical environment, detects objects, allows users to remove an object, identifies aspects like color and texture, and uses machine learning to locate and display prospective objects that fit the environment, providing realistic augmented reality previews.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional search systems rely on individual navigation and mental visualization to find prospective objects, then users can search through millions of listings, but user satisfaction decreases and resources are wasted due to mismatches between purchased objects and physical environments
Solution Approach 1:
The system creates a digital replica of the physical environment by capturing an image of the user's space, detecting objects within it, and removing selected objects to generate a modified digital environment. This digital copy allows users to visualize prospective objects in their actual physical context before purchasing, eliminating the need for mental visualization and ensuring environmental compatibility.
Solution Approach 2:
The system introduces an intermediary processing layer between the user's physical environment and the search results. This intermediary automatically detects objects in the physical space, removes selected objects from the digital representation, and uses the resulting environment as a filter for prospective objects. This mediator ensures that recommended objects naturally fit the physical environment without requiring user expertise in spatial visualization.
2Adaptability or versatility
If users manually navigate through millions of listings to find suitable objects, then comprehensive search is possible, but time consumption and user effort increase significantly
Solution Approach 1:
The system transforms the search parameters from generic keywords and categories to environment-specific visual characteristics. By analyzing the digital representation of the user's physical environment, the system automatically extracts relevant parameters such as color schemes, spatial relationships, and existing object styles. These dynamically generated parameters filter prospective objects to match the specific environment, dramatically reducing search time while maintaining comprehensive coverage of suitable options.
Solution Approach 2:
The system performs preliminary analysis of the user's physical environment before the actual object search begins. By pre-processing the environmental image, detecting objects, and creating the modified digital environment with selected objects removed, the system prepares the search criteria in advance. This preliminary action establishes the contextual framework that guides subsequent object recommendations, eliminating the need for users to manually filter through irrelevant listings.
3Productivity
If users purchase objects without visualizing them in their physical environment, then purchasing process is quick, but return rates increase due to dissatisfaction with how objects look in situ
Solution Approach 1:
The system generates augmented reality images that superimpose prospective objects onto the user's actual physical environment, creating a realistic visual copy of how the object will appear in situ. This visual copying allows users to evaluate prospective objects in their specific spatial context before purchasing, ensuring satisfaction while maintaining quick decision-making through immediate visual feedback rather than prolonged deliberation.
Solution Approach 2:
The system provides immediate visual feedback by displaying augmented reality images that show prospective objects integrated into the user's physical environment. This feedback loop allows users to quickly assess whether an object matches their space's aesthetic and functional requirements, enabling confident purchasing decisions without delays. The system iteratively refines recommendations based on user interactions with the augmented reality visualizations, accelerating the path to satisfactory purchases.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Search techniques are described that support locating and displaying prospective objects in digital images based on a removed object. A digital image, for instance, is received by a computing device as an input depicting a physical environment with various objects and displayed in a user interface. An object depicted by the digital image is removed. An aspect of the removed object is identified. A search system leverages the aspect to locate a prospective object. The prospective object is configured for display within the digital image. As a result, the digital image is displayed in the user interface as having the configured prospective object depicted in the physical environment.