The project is devoted to the feature-based description of blurred digital images acquired by a linear shift-invariant imaging system and to the recognition of such images with respect to an image database. Finding appropriate features for object description is the key to successful recognition of degraded objects. Such features must be invariant with respect to the geometric and radiometric degradations, must be able to distinguish among objects belonging to different classes and must be sufficiently stable under random noise and other factors. The imaging process is described by a convolution g(x,y)=(f*h)(x,y), where f(x,y) and g(x,y) represent the original and observed images, respectively and h(x,y) is a systems point spread function (PSF). The proposed approach consist of describing images by features which are invariant with respect to blur (that means with respect to the system PSF) and recognizing images in the feature space.