Cellular neural networks for real-time monitoring of volcanic activity

The paper introduces a new methodology for real-time monitoring of active volcanoes, which is based on efficient video processing operations implemented by means of cellular neural network (CNN) architectures. CNNs are massive parallel analog circuits with only local interconnections between the computing elements, that are programmed in an analog way to perform almost all image processing operations. The performance of CNN-based operations is reported by simulation of some dynamic image processing tasks in active volcano monitoring. The purpose of the proposed computer-based system for volcanic image processing is twofold: on-line signalling of volcanic events of interest such as lava fountains, Strombolian explosions, ash and gas emissions, etc., and real-time extraction of quantitative information which characterises the events, i.e. geometric parameters, energy involved, type of event and so on. The performance of the present version of the system is limited, in terms of processing speed, by the simulator instead of the on-chip analog CNN, which is still under development by STMicroelectronics, Hence the system can operate well only when volcanic activity is not paroxysmal. The system has been tested on images taken both on Etna and Stromboli, volcanoes located in southern Italy, but it can easily be adapted in order to work in other volcanic areas. The technique implemented for the image-processing operations, called 'CNN-ADI', was conceived for moving image processing and combines the cumulative differences model with the computational speed and versatility of CNNs, implementing a pseudo-ADI (accumulative difference image) algorithm. The advantage of using a CNN-based version of the ADI filter lies in the possibility of real-time filtering, directly on-chip of short sequences of images to distinguish between the dynamic and static elements the frames contain. The main advantages of the present work are that not only are human operators relieved of the task of visual monitoring but it is also possible to extract on-line physical parameters of volcanic events, including event classification. (C) 1999 Elsevier Science Ltd. All rights reserved.