Hand Gesture Recognition Using Dynamic Palm Displacement Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mobile devices inaccurately recognize gestures due to varying gesture deviations based on user distance and misjudge gestures when the palm moves out of the camera's visual range, leading to inaccurate control.
Innovation Solution
A gesture recognition method that dynamically calculates a displacement threshold based on palm width and orientation, using a mobile device's image capture unit to record finger coordinates and determine valid gestures, ensuring precise control by considering palm orientation, motion direction, and velocity thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed gesture displacement threshold is used for gesture recognition, then the gesture recognition process is simple, but the gesture recognition accuracy deteriorates due to varying user distances from the camera
Solution Approach 1:
The patent applies the dynamics principle by transforming the fixed displacement threshold into a dynamic threshold that adapts to different user distances. The system calculates a dynamic displacement threshold based on the detected distance between the user and the camera, allowing the gesture recognition system to maintain high accuracy across varying distances while keeping the overall process relatively simple.
Solution Approach 2:
The patent implements parameter changes by adjusting the displacement threshold parameter according to the user distance. Instead of using a constant threshold, the system modifies the threshold value based on the detected distance parameter, enabling accurate gesture recognition whether the user is close to or far from the camera.
2Area of stationary object
If the camera visual range is limited, then the device structure is compact, but gesture recognition reliability deteriorates when the palm moves out of the visual range
Solution Approach 1:
The patent applies preliminary action by tracking the palm's motion trajectory before it exits the camera's visual range. The system records the sequence of palm positions and uses this historical trajectory information to continue recognizing gestures even after the palm moves out of the current visual field, ensuring reliable gesture recognition throughout the entire gesture duration.
Solution Approach 2:
The patent uses the palm motion trajectory as an intermediary to bridge the gap between the limited camera visual range and the need for continuous gesture recognition. By storing and analyzing the trajectory data, the system can infer ongoing gestures without requiring the palm to remain continuously visible, thus maintaining reliability despite the limited visual area.
3Device complexity
If conventional gesture recognition is used without considering palm orientation, then the processing is simple, but measurement precision deteriorates due to misjudgment of gesture direction
Solution Approach 1:
The patent applies segmentation by dividing the gesture recognition process into distinct components: palm detection, orientation calculation, trajectory analysis, and gesture determination. By calculating the orientation angle of the palm relative to the screen and analyzing the motion trajectory separately, the system accurately determines gesture direction while keeping each processing step manageable and clear.
Data Source
AI summary
A gesture recognition method includes steps of (S1) capturing a current image; (S2) determining whether the image includes a hand image; (S3) when a determination result of step S2 is satisfied, determining whether the hand image is a valid gesture; (S4) when a determination result of step S3 is satisfied, the gesture recognition unit records multiple coordinate information and time information of multiple locations corresponding to multiple fingers in a finger list, and storing the finger list in the storage unit; (S5) the gesture recognition unit obtaining an orientation of the palm; (S6) the gesture recognition unit calculating a width of the palm and a dynamic displacement threshold; (S7) the gesture recognition unit determining whether to launch an action command according to the orientation of the palm, the dynamic displacement threshold, and a determination condition; (S8) when the determination result of step S7 is satisfied, launching the action command.