Optical Hand Recognition for Manual Operating Units
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems for recognizing operating intentions with joysticks in agricultural vehicles are limited by design constraints due to space requirements for sensor zones, making them unsuitable for arbitrary construction and small, finger-actuated applications, leading to potential unintended actuations.
Innovation Solution
A system utilizing an image capture unit, such as a stereoscopic camera, to monitor hand positioning and generate three-dimensional image information, which is compared to characteristic patterns in a pattern memory for intentional actuation recognition, enabling functions independently of the operating unit's design, using a 'deep learning' approach to define permitted variation ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If flat sensor zones are used for capacitive touch recognition, then operating intention recognition is enabled, but design freedom of the operating lever shape is limited and small finger-actuated elements are unsuitable
Solution Approach 1:
The patent replaces the mechanical capacitive sensor zones with an optical image capture system. Instead of using flat sensor zones that require specific contact positions, the system uses an image capture unit to detect hand and finger positions optically, enabling recognition of operating intentions on operating elements of any shape or size without being constrained by sensor zone placement requirements
Solution Approach 2:
The system creates an optical copy (image) of the hand position and compares it with stored image patterns representing intentional actuations. This allows the system to recognize operating intentions based on visual patterns rather than requiring physical contact with fixed sensor zones, thereby enabling use with various operating unit designs including small finger-actuated elements
2Reliability
If conventional sensor zones are used, then operating intention can be recognized, but unintended actuations cannot be prevented and the system is not suitable for arbitrary construction
Solution Approach 1:
The system continuously captures images of the operating unit and hand positions, compares them with stored patterns representing intentional actuations, and provides feedback to determine whether to enable or disable functions. This continuous monitoring and comparison feedback mechanism allows the system to distinguish between intentional and unintentional contact, preventing unintended actuations while working with arbitrary operating unit constructions
Solution Approach 2:
The system pre-stores multiple image patterns in a pattern memory that represent various intentional actuation scenarios. By having these patterns prepared in advance, the system can quickly compare real-time captured images against known intentional patterns, enabling reliable distinction between intentional and unintentional actuations before functions are enabled
Data Source
AI summary
A system for recognizing an operating intention at an operating unit that can be actuated manually includes an image capture unit for monitoring a capture region surrounding the operating unit for an entering hand in order to capture it in terms of a positioning of a palm of a hand or of one or a plurality of fingers in relation to the operating unit and to generate corresponding image information. An evaluation unit is provided for comparing the generated image information with at least one image pattern that is characteristic for an operating intention from a pattern memory. In the event of a correspondence that lies within a permitted variation range, a controllable function is enabled on the part of the operating unit.