Pinch Gesture Recognition via Arc Intersection Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Head-mounted display (HMD) devices for virtual reality (VR)/augmented reality (AR) face challenges in accurately and efficiently recognizing user gestures in real-time due to limited computing resources, which hinders their ability to provide seamless interactive experiences.
Innovation Solution
A gesture recognition method and device that analyzes hand images by determining a reference point, circular arc reference lines, and intersection points to assess the approaching trend of finger blocks, allowing for real-time recognition of pinch gestures without extensive data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the entire image including the hand image is captured and image recognition processing is performed on the entire image, then the gesture recognition accuracy is improved, but the computing resource consumption increases
Solution Approach 1:
The patent divides the image processing task into two stages: first capturing the entire image to locate the hand region, then extracting only the hand image for detailed gesture recognition. This segmentation reduces the amount of data requiring intensive processing while maintaining recognition accuracy.
Solution Approach 2:
The patent extracts the hand image from the entire captured image based on hand region detection, then performs gesture recognition only on this extracted hand image. This extraction approach eliminates the need to process the entire image, significantly reducing computing resource consumption while preserving gesture recognition accuracy.
2Measurement precision
If the entire image is processed for gesture recognition, then the recognition accuracy is improved, but the processing time increases
Solution Approach 1:
The patent segments the processing workflow into rapid hand region detection followed by focused gesture recognition on extracted hand images. This time-efficient segmentation enables real-time gesture recognition while maintaining accuracy.
Solution Approach 2:
By extracting only the hand image portion from the entire image for gesture recognition processing, the patent reduces processing time significantly while preserving the ability to accurately recognize gestures, enabling real-time interaction.
3Weight of stationary object
If the device body is made compact and lightweight, then the user comfort is improved, but the computing processing capability is reduced
Solution Approach 1:
The patent extracts and processes only the essential hand image data for gesture recognition, reducing the computational burden on the HMD device. This approach enables compact and lightweight device design while maintaining sufficient processing capability for real-time gesture recognition.
Solution Approach 2:
Instead of processing the entire image, the patent applies partial action by processing only the extracted hand image portion, which reduces computational requirements and enables lightweight device design while achieving the necessary processing capability for gesture recognition.
Data Source
AI summary
A gesture recognition method and a gesture recognition device are provided. The gesture recognition method includes the steps of: obtaining a hand image including a gesture graphic; determining a reference point in the gesture graphic; determining circular arc reference lines by using the reference point as a center; determining intersection points of each of the circular arc reference lines intersecting with a boundary of the gesture graphic; determining whether at least two finger blocks of a plurality of finger blocks of the gesture graphic conform to an approaching trend according to the circular arc reference lines and the intersection points, and determining whether the at least two finger blocks in a selected range of the gesture graphic forms a continuous graphic block; and when the at least two finger blocks of the gesture graphic conform to the approaching trend and form the continuous graphic block, determining the hand image as a hand pinch image.