User-Specific Gesture Recognition Using 9-Axis Motion Profiles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing gesture recognition technologies are limited in everyday use due to recognition errors, environmental dependence, and user inconvenience, and lack of methods to enhance accuracy using multi-axis sensors on wearable devices.
Innovation Solution
A method for generating a user-specific gesture profile based on 3D motion data using a 9-axis sensor, including an accelerometer, gyroscope, and magnetometer, to accurately recognize gestures through reference motion analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image processing method is used for gesture recognition, then gesture recognition can be performed, but recognition errors are frequent and algorithm needs to be very sophisticated
Solution Approach 1:
The patent replaces image processing (optical system) with motion sensor data processing (mechanical sensing system). Instead of capturing images and processing visual information, the system uses accelerometers, gyroscopes, and magnetometers to directly measure motion parameters, thereby avoiding the complexity of image processing algorithms while improving recognition accuracy.
Solution Approach 2:
The patent introduces motion sensor data as an intermediary between the gesture and the recognition system. Rather than directly processing images or visual information, the system uses sensor data (acceleration, angular velocity, magnetic field) as a mediator to capture and interpret gesture movements, simplifying the recognition process.
2Measurement precision
If landmark or special gear is attached to increase recognition accuracy, then gesture recognition accuracy is improved, but it becomes too cumbersome and inconvenient for user
Solution Approach 1:
The patent enables the user's body to serve itself for gesture recognition. By placing motion sensors on the user's body (e.g., in a smartphone held in hand or worn on wrist), the system uses the user's own motion to perform recognition without requiring external landmarks, special clothing, or additional accessories. The user's natural movement becomes the recognition medium.
Solution Approach 2:
The patent makes the motion sensor system universal by using common devices (smartphones, wearables) that users already possess. The same sensor platform serves multiple functions: daily communication, navigation, and gesture recognition, eliminating the need for specialized equipment while maintaining convenience.
3Productivity
If only one or two types of sensor data are used for gesture recognition, then processing is simplified, but gesture recognition rate is not effectively increased
Solution Approach 1:
The patent merges multiple types of sensor data (acceleration from accelerometer, angular velocity from gyroscope, magnetic field information from magnetometer) into a unified gesture recognition system. By combining these complementary data sources, the system achieves higher recognition rates than using any single sensor type alone, while the integrated processing approach manages complexity through coordinated multi-sensor fusion.
Solution Approach 2:
The patent creates a composite sensing system that combines different sensor technologies (accelerometer, gyroscope, magnetometer) into a unified 9-axis sensor platform. This composite approach leverages the strengths of each sensor type to achieve superior gesture recognition performance that cannot be obtained with单一 sensor types.
Data Source
AI summary
A method by which an electronic device outputs a control signal based on a motion of a user may comprise the steps of: when an operation mode of the electronic device is a gesture recording mode, generating a gesture profile of the user corresponding to a reference motion as the user performs a recording motion of mimicking the reference motion at least once; and when the operation mode is a gesture recognition mode, determining whether a recognition motion of the user matches the reference motion.