We describe a nonparametric approach of dynamic thermography to the detection of buried antipersonnel (AP) mines. Dynamic thermography consist of processing temporal sequences of IR infrared images taken from the same scene submitted to either artificial of natural temperature variations. The aim is to obtain an image segmentation where mine and soil can be discriminated due to the different time evolution of their thermal properties. The proposed approach is rooted in a clustering stage performed by a chaotic neural network and provides the correct classification by analyzing very short image seqences, thus enabling a fast acquisition time. The effectiveness of the method is demonstrated on image sequences of plastic AP mines taken from realistic mine fields.
Chaotic neural network clustering: an application to landmine detection by dynamic infrared imaging
C Marangi;G Satalino;
2001
Abstract
We describe a nonparametric approach of dynamic thermography to the detection of buried antipersonnel (AP) mines. Dynamic thermography consist of processing temporal sequences of IR infrared images taken from the same scene submitted to either artificial of natural temperature variations. The aim is to obtain an image segmentation where mine and soil can be discriminated due to the different time evolution of their thermal properties. The proposed approach is rooted in a clustering stage performed by a chaotic neural network and provides the correct classification by analyzing very short image seqences, thus enabling a fast acquisition time. The effectiveness of the method is demonstrated on image sequences of plastic AP mines taken from realistic mine fields.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
