Image Processing Unit Virtual Object Placement Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies fail to accurately combine virtual objects with captured images based on the state or type of objects shown, leading to inappropriate processing and mismatched virtual object placement, such as excess clothing parts being difficult to erase or accessories not matching the user's attire.
Innovation Solution
An image processing apparatus with an image processing unit that determines virtual objects to be combined with a captured image based on the state or type of objects shown, using detection units to identify features and adjust virtual object placement accordingly, such as erasing excess clothing parts or modifying accessories to match the user's attire.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If virtual objects are combined with captured images using existing technologies, then image processing can be performed, but the virtual object placement and object processing are inaccurate and do not reflect the actual state or type of objects shown in the captured image
Solution Approach 1:
The system performs preliminary detection and recognition of objects, their states, and types in the captured image before combining virtual objects. This preliminary analysis enables accurate determination of virtual object placement and appropriate selection/modification based on the detected object characteristics, resolving the contradiction between precision and adaptability
Solution Approach 2:
The system uses feedback from object detection results to adjust virtual object selection and placement. By continuously analyzing the detected object state and type, and using this information to refine virtual object combination, the system achieves both accurate placement and adaptability to different object characteristics
2Ease of operation
If clothes that a user already wore come out of the clothes, then the captured image shows the excess part, but it is difficult to automatically erase the part that came out
Solution Approach 1:
The system performs preliminary detection of excess clothing parts that come out of other clothes before the erasing operation. By identifying these parts in advance based on object state analysis, the system can then automatically and accurately erase only the detected excess portions, resolving the contradiction between ease of operation and manufacturing precision
Solution Approach 2:
The system extracts and identifies the specific excess clothing parts that need to be removed from the captured image. By separating the detection of excess parts from the general object recognition, the system enables precise targeting of only the portions that need erasing, achieving both automation and accuracy
3Adaptability or versatility
If accessories are combined with captured images without considering the user's attire, then the processing is simple, but the accessory does not match the user's attire
Solution Approach 1:
The system performs preliminary detection and classification of the user's attire (clothes type, style, color) before selecting accessories. This preliminary analysis of object type and state enables the system to automatically determine appropriate accessories that match the detected attire characteristics, resolving the contradiction between adaptability and device complexity
Solution Approach 2:
The system uses the detected attire information to automatically select and combine appropriate accessories without requiring manual input or complex configuration. By making the system self-sufficient in determining accessory matching based on detected object characteristics, the complexity is managed internally while maintaining high adaptability
Data Source
AI summary
There is provided an image processing apparatus including an image processing unit which combines a virtual object with a captured image. The image processing unit determines the virtual object based on a state or a type of an object shown in the captured image.