In this work, a heterogeneity feature, calculable from SAR images on a per-pixel basis, but relying on global image statistics, is described and discussed. Starting from the multiplicative speckle and texture models relating the amount of texture and speckle to the local mean and variance at every pixel, such a feature is rigorously derived from Shannon's information theory as the conditional information of local standard deviation to local mean. Thanks to robust statistical estimation, it is very little sensitive to the noise affecting SAR data and thus capable of capturing subtle variations of texture, whenever they are embedded in a heavy speckle. Experimental results carried out on a variety of SAR images with different degrees of noisiness demonstrate that the proposed feature is likely to be useful for a variety of automated segmentation and classification tasks.
An information-theoretic sar heterogeneity feature / Aiazzi, Bruno; Alparone, Luciano; Baronti, Stefano. - ELETTRONICO. - (2004), pp. 17-24. (Intervento presentato al convegno ESA-EUSC 2004: Theory and Applications of Knowledge-Driven Image Information Mining with Focus on Earth Observation tenutosi a Madrid, esp nel 17 - 18 March 2004).
An information-theoretic sar heterogeneity feature
ALPARONE, LUCIANO;
2004
Abstract
In this work, a heterogeneity feature, calculable from SAR images on a per-pixel basis, but relying on global image statistics, is described and discussed. Starting from the multiplicative speckle and texture models relating the amount of texture and speckle to the local mean and variance at every pixel, such a feature is rigorously derived from Shannon's information theory as the conditional information of local standard deviation to local mean. Thanks to robust statistical estimation, it is very little sensitive to the noise affecting SAR data and thus capable of capturing subtle variations of texture, whenever they are embedded in a heavy speckle. Experimental results carried out on a variety of SAR images with different degrees of noisiness demonstrate that the proposed feature is likely to be useful for a variety of automated segmentation and classification tasks.I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.