Work packages

The project is broken into the following work packages:

Work package 1: Visual Neural Systems Modelling for near range emergent collision detection

1.1 To model invertebrate visual neural systems for collision detection

1.2 To enhance collision detection by integrating multiple invertebrate visual neural systems

1.3 To integrate both invertebrate and vertebrate visual system models

 

Work package 2: Brain-inspired vision systems for long-range hazard perception

2.1 To develop efficient long range hazard perception methods

2.2 To develop efficient signpost text and road markings recognition methods

2.3 To develop long range hazard perception methods coping with low light conditions

 

Work package 3: Visual computation of multi-modality sensor data for collision detection

3.1 To select, implement thermal image sensor, prepare pre-processing algorithms

3.2 To identify bio-inspired neural system models for thermal image processing

3.3 To explore the robustness of the thermal image based collision detection in extreme conditions

 

Work package 4: Systems integration, miniaturisation, verification and demonstration

4.1 To build up a database for verification

4.2 To integrate and implement preliminary near range imminent collision and long range hazard perception systems

4.3 To identify circuit implementations, optimal sensory/processing integration for hardware miniaturization

4.4 To build a demonstrator system for the collision avoidance

Ultra-layered perception with brain-inspired information processing for vehicle collision avoidance