Community Research and Development Information Service - CORDIS


This paper describes a computer vision system for analysing air flows. The system uses a trace gas to visualise the air flow and to determine the air flow by analysing the trace gas and its movement via computer vision and neural network techniques. The analysis process is based on three major stages: first, an image pair series is taken from the traced air flow; second, the 3-D shapes of trace gas objects at each image pair instant are recovered; third, the air flow velocity field is calculated via the self-organising map from the recovered 3-D surface models of the trace gas objects. The system can be used to measure air flow profiles, to determine air flow velocities and volume fluxes, and to estimate how small particles (e.g. dust) move within the flow.

Additional information

Authors: HEIKKONEN J, JRC Ispra (IT)
Bibliographic Reference: Paper presented: International Conference on Engineering Applications of Neural Networks, Espoo (FI), August 21-23, 1995
Availability: Available from (1) as Paper EN 39236 ORA
Record Number: 199511236 / Last updated on: 1995-10-13
Original language: en
Available languages: en