A model for space-variant two-dimensional filtering of noisy images is discussed. The signal recording process usually introduces several perturbations, whose overall effect on the detected image is called blur. It mainly depends on the low-pass filter behavior of the measurement equipment, and on typical aberration of the optical components of the imaging system. The processing of real images requires algorithms where a rapid switching of the filter characteristics is allowed. A nonstationary state-space representation is obtained, starting from the smoothness, stochastic, and inhomogeneity assumptions. The obtained model is space varying according to the presence of image edges and in this way the edge-defocusing phenomenon is greatly reduced. The synthesized image has been chosen because it contains sharp edges, while the real image has been chosen to evaluate the filter performance on real data. The adaptive behavior of the proposed restoration method is obtained by including the information on edge locations into the image model.
Space-variant two-dimensional filtering of noisy images / DE SANTIS, Alberto; Alfredo, Germani; Leopoldo, Jetto. - STAMPA. - 119:C(2001), pp. 267-318. [10.1016/s1076-5670(01)80089-6].
Space-variant two-dimensional filtering of noisy images
DE SANTIS, Alberto;
2001
Abstract
A model for space-variant two-dimensional filtering of noisy images is discussed. The signal recording process usually introduces several perturbations, whose overall effect on the detected image is called blur. It mainly depends on the low-pass filter behavior of the measurement equipment, and on typical aberration of the optical components of the imaging system. The processing of real images requires algorithms where a rapid switching of the filter characteristics is allowed. A nonstationary state-space representation is obtained, starting from the smoothness, stochastic, and inhomogeneity assumptions. The obtained model is space varying according to the presence of image edges and in this way the edge-defocusing phenomenon is greatly reduced. The synthesized image has been chosen because it contains sharp edges, while the real image has been chosen to evaluate the filter performance on real data. The adaptive behavior of the proposed restoration method is obtained by including the information on edge locations into the image model.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.