stereoscopic camera for rail

client

Rail

role

Camera, Linux, GPS, OTA

year

2024

Stereoscopic Camera for Vegetation Detection in Cab-Based Installation

This project involved the development and deployment of a stereoscopic camera system for vegetation detection, installed indoors at the front of a cab. The system was designed to capture high-resolution 3D imagery for detecting and analysing vegetation along the vehicle's path. The camera hardware was integrated with third-party software, though the software development was not part of the scope. The project followed these key stages:

1. Requirements Analysis & Design

The initial phase involved analysing the specific requirements for vegetation detection from a moving vehicle. The system needed to capture detailed 3D imagery while mounted inside the cab, focusing on detecting vegetation close to the vehicle’s path. Key considerations included minimising the impact of vehicle movement and varying indoor lighting conditions, as well as ensuring that the camera was rugged enough for long-term use in the field.

2. Hardware Development

A custom stereoscopic camera system was designed, equipped with dual lenses to capture synchronized images for creating detailed 3D models of the surrounding vegetation. The camera was optimised for installation inside the cab, ensuring stable and high-resolution image capture despite the vehicle’s motion. Special attention was given to mounting mechanisms and calibration to achieve accurate depth perception, necessary for vegetation detection.

3. Platform Delivery & Installation Assistance

The camera system was installed at the front of the vehicle’s cab, positioned to provide a clear view of the path ahead. Assistance was provided during installation to ensure proper calibration and alignment of the stereoscopic lenses, allowing for accurate 3D imagery capture. The hardware was integrated with a third-party software platform responsible for processing the 3D data to analyse vegetation, but the software deployment was managed separately.

‍

4. Prototyping & Testing

Prototypes of the camera were tested in indoor environments and field conditions, focusing on the system's ability to capture precise 3D imagery while the vehicle was in motion. Testing scenarios included various lighting conditions and vehicle speeds to ensure the hardware performed reliably in real-world operation. Adjustments were made based on test results to optimise the image stability and depth accuracy during movement.

5. Deployment & Operation

The full stereoscopic camera system was deployed in multiple vehicles, capturing continuous 3D imagery of vegetation as the cab moved along its route. The hardware was designed to work seamlessly with third-party software, which processed the captured images for vegetation detection and analysis. The system provided consistent data despite the movement of the vehicle, ensuring accurate vegetation mapping.

6. Results & Future Enhancements

The hardware successfully captured high-quality 3D imagery, providing the depth data necessary for detailed vegetation analysis. By integrating with the third-party software, the system delivered real-time vegetation detection while the vehicle was in motion. The project demonstrated the feasibility of cab-based stereoscopic cameras for environmental monitoring, with potential for future improvements in sensor stability and depth accuracy.