Showing papers by "Alan C. Bovik published in 1987"

The Effect of Median Filtering on Edge Estimation and Detection

Abstract: In this paper we consider the effect of median prefiltering on the subsequent estimation and detection of edges in digital images. Where possible, a quantitative statistical comparison is made for a number of filters defined with two-dimensional geometries; in some cases one-dimensional analyses are required to illustrate certain points. Noise images prefiltered by median filters defined with a variety of windowing geometries are used to support the analysis, and it is found that median prefiltering improves the performance of both thresholding and zero-crossing based edge detectors.

Streaking in median filtered images

Abstract: This paper presents a probabilistic analysis of the streaking or blotching effect commonly observed in median filtered signals in both one and two dimensions. The effcts are identified as runs of equal or nearly equal values which create visual impressions that have no visual correlate. For one-dimensional discrete iid random signals with continuous input probability densities, the probability of a streak of length L occurring is computed and shown to be independent of the input probability distribution. Expressions for the first and second moments of the streak length are also derived, and certain asymptotic results are given. As the analysis and definition of the analogous effect in two dimensions is less tractable, the probability that medians taken from distinct overlapping windows will take the same value is derived for various filter geometries. The analytic results are supported by examples using both one- and two-dimensional signals.

Texture segmentation using a class of narrowband filters

Abstract: A class of 2D filters is proposed for segmenting visible images into regions of uniform texture. The filters used, known as Gabor filters, are optimal in several senses: they have tunable orientation bandwidths, they can be defined to operate over a range of spatial frequency channels, and they obey the uncertainty principle in two dimensions. The filters are interpreted as transforming the image into a modulated narrowband signal whose envelope coincides with the textured region to which the filter is tuned. Moreover, the receptive fields of neurons in the visual cortex are known to have shapes that approximate 2D Gabor filters, whose purpose has been uncertain. We suggest that they may play an important role in texture segmentation/surface perception. The technique is demonstrated using a variety of natural and synthetic textures.

Spectral properties of moving L-estimates of independent data

Abstract: A derivation of the joint probability distribution and mass functions of order statistics coming from overlapping samples is presented. The general formulation allows for samples of any size overlapping (coinciding) in any number of observed values ranging from zero to the number of observations in the smaller sample. These expressions are used to compute the autocovariance function of a moving L -estimate (linear combination of order statistics) of a sequence of independent, identically distributed second-order random variables, under a variety of assumptions on the parent distribution. The associated variance spectral density is also computed for several filters of interest, including median filters, and inner and outer trimmed mean filters.

Experiments with a theory of visual texture segmentation using modulation/demodulation processes

Abstract: The receptive field profiles of certain simple cells in the visual cortex have recently been shown to approximate the 2-D Gabor functions, which are Gaussian-modulated sinusoidal gratings in the space domain, and shifted Gaussians in the frequency domain (Marcelja 1980; Daugman 1980,1985). The 2-D Gabor functions minimize the space-frequency uncertainty principle, and hence can be defined with narrow frequency and orientation responses while maintaining spatial localization. Physiological evidence indicates that the Gabor receptive fields are often arranged in quadrature pairs (Pollen and Ronner 1981) allowing a novel interpretation of early visual processing whereby highly frequency- or orientation-specific information can be recovered from the ampUtude or phase envelopes of a set of Gabor-filtered images. In particular, this leads to an interpretation of texture as a surface code from which physical information can be derived via appropriate filtering/ demodulation processes.

Automated 3-D analysis of stereo-microscopic images

Abstract: In this paper, we discuss the 3-D analysis of microscopic-scale biological objects using stereo vision techniques. First, we describe a number of observations on problems encountered in stereoscopic matching of images containing microscopic-scale biological objects. An algorithm is also presented for matching, which is based on the disparity gradient constraint and utilizes image intensities as the matching input. A number of examples are given illustrating the performance of the algorithm on stereo-microscopic images. While we have obtained good results for images containing opaque objects, we have still been unable to obtain acceptable results for objects which largely exhibit transparency. We also explain the reason why the matching is difficult in transparent surfaces.