🚀 go-pugleaf

One of the most successful ESPRIT projects funded by the EEC developed
a control system for a glass melting furnace which uses images from a
camera installed inside the furnace for control and report purposes.
The images are processed and pattern recognition algorithms help to
determine the quality of the flame, eventually helping to determine
the cause for a malfunction. Other information provided for an algorithm
that optimizes the parameters of PID controllers used in the low level
control system are parameters relative to the flame, such as its length
and width. The images also contain information about the presence of
unmelted glass parts.

The prototype was already tested with significant savings of fuel. Before,
the information was seldom used by an operator to check the flame.
Due to this success, the funding of the project is to be continued.

I have here two references about the vision part of this work (also involving
control optimization and expert control). For more information, please
send email to
	jasv%kappa.ist.utl.pt@inesc.pt

1. J. P. Costeira, J. P. Fernandes, M. V. Heitor, J. Sentieiro, J. Simo~es,
	J. A. Victor, "A Vision System for Analysis and Classification of
	Industrial Flames", 6th CIM-EUROPE Conf., Lisbon, May 1990

2. J. Victor, J. Costeira, J. Tome', J. Sentieiro, "A Computer Vision System
	for the Characterization and Classification of Flames in Glass 
	Furnaces", Proc. of the 1991 IEEE IAS Meeting, Dearborn, Michigan,
	Oct 91. To be published in IEEE Trans. on Ind. Appl.

I forgot to mention that the name of the ESPRIT project is AIMBURN and
the vision system was developed by two groups of Lisbon Technical University,
the ISR (Institute of Systems and Robotics) and CTAMFUL (Center for
Applied Thermodynamics and Fluid Mechanics of Lisbon University).

Pedro Lima
lima@ral.rpi.edu