Rail Vehicle Stereo Vision for Obstacle Depth Perception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Rail vehicles operating autonomously in traffic spaces with other road users face challenges in evasive maneuvers due to their fixed route guidance and limited ability to assess obstacles, leading to potential collisions and restricted operation.
Innovation Solution
A rail vehicle equipped with an image generation system comprising at least four image generation devices forming two stereo pairs with different viewing angles, allowing for reliable depth information calculation and continued operation even with faulty devices, along with redundant processor units and transmission systems for enhanced reliability and remote monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If a rail vehicle uses fixed route guidance and envelope-based space requirements, then the vehicle can operate autonomously on tracks, but the vehicle cannot perform evasive maneuvers or circumnavigate obstacles when collisions are detected
Solution Approach 1:
The image generation system is divided into multiple independent image generation devices (at least four devices) that can be segmented into different stereo pairs. This segmentation allows the system to maintain autonomous operation while providing multiple independent viewing angles for obstacle detection, enabling better adaptability without requiring physical evasive maneuvers.
Solution Approach 2:
The system transitions from two-dimensional image capture to three-dimensional depth perception by implementing multiple stereo pairs with different baselines. This dimensional enhancement allows the vehicle to assess obstacle depth and position more accurately, compensating for the inability to perform lateral evasive maneuvers.
2Device complexity
If a rail vehicle uses a single stereo pair for depth perception, then the device complexity is reduced, but the system becomes vulnerable to failures and cannot maintain operation when image generation devices fail
Solution Approach 1:
Different image generation devices are assigned different functions and viewing angles. The system creates at least two stereo pairs from four devices, where each stereo pair can independently provide depth information. This local differentiation ensures that if one device fails, the remaining devices can still form functional stereo pairs and maintain system operation.
Solution Approach 2:
The system is designed with redundant image generation devices and multiple stereo pair configurations before any failure occurs. This beforehand cushioning ensures that the rail vehicle can continue autonomous operation even when devices fail, as the remaining devices can still provide sufficient depth perception capability.
3Measurement precision
If a rail vehicle uses multiple image generation devices with different viewing angles, then depth information accuracy and system reliability are improved, but the device complexity and number of components increase
Solution Approach 1:
Each image generation device serves multiple functions: it contributes to at least one stereo pair and can potentially participate in multiple stereo pairs depending on configuration. The four devices work together to provide both depth perception and redundancy, making the system multi-functional without requiring separate dedicated sensors for each function.
Solution Approach 2:
The system merges the functions of multiple image generation devices into a unified evaluation system that processes data from all devices. By combining the capabilities of four devices into two stereo pairs with different baselines, the system achieves enhanced measurement precision while managing complexity through integrated processing.
4Reliability
If a rail vehicle requires sufficient space along the route according to its envelope, then the vehicle maintains structural integrity and operational safety, but the vehicle length and clearance requirements increase, making it more difficult to capture the relevant outside space
Solution Approach 1:
The system transitions from two-dimensional envelope-based safety assessment to three-dimensional depth perception by implementing stereo vision with multiple baseline distances. This allows accurate capture of obstacle depth information within the vehicle's required clearance space, eliminating the need for excessive lateral clearance while maintaining operational safety.
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
The system ensures reliable autonomous operation by providing robust image capture and processing capabilities, enabling continued safe operation even with obstacles and maintaining safety standards despite device failures, and allows for remote monitoring and control.
Implementation Method 1
image generation devices, such as digital cameras... to generate two-dimensional images of the space outside the rail vehicle
Implementation Method 2
a first and a second of the four image generation devices form a first stereo pair... a third and a fourth of the four image generation devices form a second stereo pair... for capturing a first and a second shared portion of the space from different viewing angles
Data Source
AI summary
The invention relates to a rail vehicle including an image generation system for capturing a space outside the rail vehicle. The image generation system includes four image generation devices, each of which, during operation of the image generation system, generates two-dimensional images of the space. A first and a second of the four image generation devices are disposed at a first distance from one another on the rail vehicle and form a first stereo pair, which captures a first shared portion of the space from different viewing angles. A third and a fourth of the four image generation devices are disposed at a second distance from one another on the rail vehicle and form a second stereo pair, which detects a second shared portion of the space from different viewing angles. The first shared portion of the space and the second shared portion of the space have a shared spatial region.