TOF Gesture Recognition Using Palm Depth Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Gesture recognition using Time-of-Flight (TOF) cameras in mobile platforms faces challenges due to low resolution depth images and low data acquisition frame rates, resulting in suboptimal accuracy for gesture recognition.
Innovation Solution
A method and apparatus that acquire depth images using a TOF camera, calibrate the camera to obtain three-dimensional point clouds, and categorize these point clouds to accurately extract and recognize hand gestures, even with low resolution and frame rate limitations, by determining point sets indicating the palm and analyzing their distribution characteristics to identify gestures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If TOF camera is used for gesture recognition, then cost is reduced and miniaturization is easier, but resolution and frame rate are low resulting in suboptimal accuracy
Solution Approach 1:
The patent segments the depth image into multiple regions including foreground region, background region, and edge region. By separately processing these regions and focusing computational resources on the foreground region where the hand is located, the system achieves accurate gesture recognition despite the low resolution inherent to TOF cameras.
Solution Approach 2:
The patent changes the parameter of region weighting by assigning different weights to different regions of the depth image. The foreground region is given higher weight while background and edge regions receive lower weights. This parameter change allows the system to extract meaningful gesture information from low-resolution TOF data by concentrating analysis on the most relevant areas.
2Device complexity
If TOF camera is used for gesture recognition, then device complexity is reduced, but data acquisition frame rate is low resulting in suboptimal accuracy
Solution Approach 1:
The patent extracts only the essential information from the low frame rate depth images by identifying and isolating the foreground region containing the hand. Instead of processing entire frames at high frequency, the system extracts hand position and gesture information from each available frame, making effective use of the limited data acquisition rate while maintaining simple TOF camera hardware.
3Volume of moving object
If low resolution depth image is used, then TOF camera can be miniaturized, but palm extraction accuracy is reduced
Solution Approach 1:
The patent introduces a regional dimension by dividing the depth image into foreground, background, and edge regions. This dimensional segmentation allows the system to process low-resolution TOF data effectively by focusing on the foreground region where palm information is located, thereby achieving accurate palm extraction despite the small camera size and low resolution.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate extraction of palm position and motion trajectory, improving gesture recognition accuracy and reducing computational resources, thereby enhancing the control of mobile platforms like UAVs with enriched control methods.
Implementation Method 1
Time-of-Flight (TOF) cameras
Data Source
AI summary
The present disclosure provides a gesture recognition method. The method includes the following steps: acquiring a depth image of a user; determining a point set of a two-dimensional image indicating a palm based on a depth information of the depth image; and, determining a gesture based on the point set.