Mobile Device Gesture Recognition Using Multi-Stage Sensor Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile computing devices face challenges in accurately recognizing and initiating actions based on user-induced physical movements, often leading to unintentional actions due to noise from hand tremors or spurious movements.
Innovation Solution
The device employs a combination of accelerometer, gyroscope, and magnetometer sensors to detect translational and rotational movements, using weighted sensor data and a low-pass filter to differentiate between intended and unintended movements, thereby initiating predefined actions only when specific criteria are met.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion sensors are used to detect physical movements for initiating actions, then user gesture recognition capability is improved, but false detection from hand tremors and spurious movements increases
Solution Approach 1:
The patent segments the gesture recognition process into multiple independent evaluation stages: (1) detecting candidate gestures from sensor data, (2) evaluating each candidate against multiple criteria including motion magnitude, duration, and pattern matching, (3) filtering out false positives by requiring satisfaction of all criteria thresholds. This multi-stage segmentation allows the system to achieve high recognition accuracy while maintaining reliability by progressively eliminating false detections.
Solution Approach 2:
The system implements feedback mechanisms where the evaluation of each gesture candidate incorporates real-time sensor data streaming and continuous threshold checking. The system monitors gesture progression over time, adjusting evaluation based on whether the gesture is completing or being aborted, and provides feedback by either confirming the gesture to trigger the action or rejecting it to prevent false initiation. This dynamic feedback loop ensures both accurate recognition and high reliability.
2Reliability
If multiple sensor criteria are evaluated to reduce false detections, then action initiation reliability is improved, but gesture recognition speed decreases
Solution Approach 1:
The patent applies preliminary action by pre-establishing all evaluation criteria thresholds and gesture patterns before runtime. The system pre-configures acceptable motion magnitudes, duration ranges, and pattern templates for different gesture types. During actual gesture detection, the system only needs to check against these pre-defined criteria rather than performing complex real-time analysis, significantly reducing processing time while maintaining high reliability through comprehensive criterion evaluation.
3Ease of operation
If motion threshold is set low to capture subtle gestures, then ease of operation is improved, but unintentional actions from noise increase
Solution Approach 1:
The patent moves the evaluation from a single-dimensional threshold check to multi-dimensional assessment by considering multiple gesture attributes simultaneously: motion magnitude, duration, acceleration patterns, and temporal progression. Instead of relying solely on a low motion threshold that would capture subtle gestures but also noise, the system evaluates the combination of multiple dimensions, allowing subtle valid gestures to be recognized while noise fails to satisfy all dimensional criteria, thus preventing false actions.
Data Source
AI summary
In one embodiment, a method includes, by a computing device, receiving sensor data from a sensor on the computing device indicating physical movement of the computing device. The method also includes determining a motion-trigger signal of the sensor data corresponding to a first characteristic of the physical movement of the computing device and a motion-confirm signal of the sensor data corresponding to a second characteristic of the physical movement of the computing device. The method further includes determining whether the motion-trigger signal includes a transition from within a pre-defined threshold band to outside of the pre-defined threshold band and whether the motion-confirm signal is within the pre-defined threshold band. The method also includes initiating a pre-defined action of the computing device when the motion-trigger signal includes the transition from within the pre-defined threshold band to outside the pre-defined threshold band and the motion-confirm signal is within the pre-defined threshold band.