Vehicle Gesture Recognition via Neural Network Video Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle interaction systems are cumbersome, requiring direct user contact or physical devices, and lack ergonomic and secure solutions for authorized access.
Innovation Solution
A method using video data captured by wide-angle cameras to recognize authorized users through predetermined gestures or postures via a neural network, allowing actuations like unlocking or locking without physical contact, using face recognition and segmentation algorithms for secure authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If physical key or key fob is used for vehicle interaction, then security is maintained, but ease of operation deteriorates due to cumbersome interaction requirements
Solution Approach 1:
The patent replaces mechanical interaction systems (physical keys, key fobs requiring direct contact) with a vision-based neural network system that captures video data and recognizes gestures/postures remotely, eliminating the need for physical devices and direct contact with the vehicle
Solution Approach 2:
The patent introduces video data as an intermediary between the user and the vehicle system. The neural network processes video data to recognize authorized users and their gestures, serving as a mediator that translates visual information into vehicle control commands without requiring physical contact
2Ease of operation
If direct contact with vehicle is required for interaction, then security is improved, but ease of operation deteriorates
Solution Approach 1:
The patent replaces mechanical contact-based authentication with a vision-based system that uses video data and neural network recognition to authenticate users remotely through gesture and posture analysis, maintaining security while improving ease of operation
Solution Approach 2:
The patent changes the authentication parameters from physical contact requirements to visual gesture recognition. The neural network analyzes video data to detect specific gestures and postures, transforming the authentication mechanism from tactile to visual while maintaining security reliability
3Ease of operation
If physical devices are used for vehicle access, then reliability is maintained, but ease of operation deteriorates due to cumbersome interaction
Solution Approach 1:
The patent creates a universal interaction system where the neural network can recognize multiple types of gestures and postures for different vehicle operations (unlocking, locking, opening doors). This single vision-based system replaces multiple physical devices and interaction methods, providing both ease of operation and adaptability
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The disclosure notably relates to a method of user-interaction with a vehicle. The method comprises, by a user, approaching the vehicle from the outside. The method also comprises, by the vehicle, capturing video data of the user. The method also comprises, by the user, while being outside the vehicle and while the vehicle captures the video data of the user, performing a predetermined gesture or taking a predetermined posture. The method also comprises by the vehicle, determining that the user is an authorized user and, by a respective neural network function and based on at least a respective portion of the video data, recognizing the predetermined gesture or the predetermined posture. The method also comprises if the user is an authorized user, performing one or more corresponding actuations.