Multi-pixel anomaly detection in multi-temporal thermography

Anomaly detection in image processing has been used for a variety of application. The main objective is finding a pixel that differs significantly from the background. We assume no a-priori information on the data. For multi-temporal thermography, each pixel contains samples of the temperature for different times. In this paper, we will present a new method for anomaly detection that combines the spatial coherence of our images with the anomaly detection RX algorithm. Throughout this article, we compare the results of our algorithm, with the results of RX algorithm, and its variations. This article applies multi-spectral algorithms to multi-temporal thermography. A novel algorithm is suggested. This algorithm does not assume the target´s size is known nor its orientation, and works with any target size, and orientation. The algorithm can be run in parallel to improve run time. The suggested algorithm is applied to simulated and real data.