Rotation Detection via Facial Key Points in Smart Displays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for human presence detection and face recognition in smart applications face performance issues when image rotation is needed, and relying on sensors like accelerometers or gyroscopes may not provide beneficial results for these applications.
Innovation Solution
A judgment system comprising a feature acquisition module and a judgment module that analyzes key point coordinates and face box size in an image to determine if a rotation signal should be sent, allowing the system to adjust the orientation of screen-displayed content accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor-based rotation detection is used, then rotation information can be obtained, but the judgment result may not be beneficial for smart applications and time cost increases
Solution Approach 1:
The patent replaces sensor-based mechanical rotation detection with image processing-based rotation detection. By analyzing the orientation of human body key points (such as head, shoulders, hips) in images, the system determines rotation state without relying on accelerometers or gyroscopes, thereby eliminating sensor data processing time and improving detection accuracy for smart applications.
Solution Approach 2:
The patent introduces image analysis as an intermediary method between the camera and rotation detection. Instead of directly using sensor data, the system uses images of human bodies and analyzes the spatial relationships and orientations of key points to infer rotation state, providing a more accurate and application-specific rotation detection mechanism.
2Reliability
If sensor-based rotation detection is used, then rotation information can be obtained, but the judgment result may not be beneficial for smart applications
Solution Approach 1:
The patent applies local quality analysis by focusing on specific local features of the human body (head, shoulders, hips, legs) and their spatial relationships. By analyzing the orientation and position of these specific body parts in the image, the system can accurately determine rotation state that is relevant to smart applications like human presence detection and face recognition, rather than using generic sensor data.
Solution Approach 2:
The patent creates a virtual model of human body orientation by extracting key point coordinates and spatial relationships from images. This virtual representation of body orientation serves as a copy that can be processed to determine rotation state, providing application-specific rotation information without relying on sensor data.
3Loss of time
If image processing for rotation detection is implemented, then calculation time is reduced, but device complexity increases
Solution Approach 1:
The patent segments the human body into distinct key points (head, shoulders, hips, legs) and analyzes each segment's orientation separately. This segmentation approach simplifies the image processing by breaking down the complex task of rotation detection into smaller, manageable analyses of individual body parts and their spatial relationships, reducing overall computational complexity.
Solution Approach 2:
The patent uses partial action by analyzing only the necessary key points and spatial relationships needed for rotation detection, rather than processing the entire image in detail. By focusing on specific body parts and their orientations, the system achieves sufficient accuracy for smart applications with reduced computational effort.
Data Source
AI summary
A judgment system, an electronic system, a judgment method, and a display method are provided. The judgment method includes: receiving an image by a feature acquisition module and obtaining a first key point coordinate, a second key point coordinate, and a size of a face box of a user by the feature acquisition module based on the image; and performing following steps by a judgment module: obtaining a judgment value based on an ordinate of the first key point coordinate, an ordinate of the second key point coordinate, and a size of the face box; and sending a rotation signal in response to that the judgment value satisfies a rotation condition.