Boaz J. Super

Edge co-occurrence in natural images predicts contour grouping performance.

The human brain manages to correctly interpret almost every visual image it receives from the environment. Underlying this ability are contour grouping mechanisms that appropriately link local edge elements into global contours. Although a general view of how the brain achieves effective contour grouping has emerged, there have been a number of different specific proposals and few successes at quantitatively predicting performance. These previous proposals have been developed largely by intuition and computational trial and error. A more principled approach is to begin with an examination of the statistical properties of contours that exist in natural images, because it is these statistics that drove the evolution of the grouping mechanisms. Here we report measurements of both absolute and Bayesian edge co-occurrence statistics in natural images, as well as human performance for detecting natural-shaped contours in complex backgrounds. We find that contour detection performance is quantitatively predicted by a local grouping rule derived directly from the co-occurrence statistics, in combination with a very simple integration rule (a transitivity rule) that links the locally grouped contour elements into longer contours.

Comparison of five color models in skin pixel classification

Detection of skin in video is an important component of systems for detecting, recognizing, and tracking faces and hands. Different skin detection methods have used different color spaces. This paper presents a comparative evaluation of pixel classification performance of two skin detection methods in five color spaces. The skin detection methods used in this paper are color-histogram based approaches that are intended to work with a wide variety of individuals, lighting conditions, and skin tones. One is the widely-used lookup table method, the other makes use of Bayesian decision theory. Two types of enhancements, based on spatial and texture analyses, are also evaluated.

Shape from texture using local spectral moments

Presents a non-feature-based solution to the problem of computing the shape of curved surfaces from texture information. First, the use of local spatial-frequency spectra and their moments to describe texture is discussed and motivated. A new, more accurate method for measuring the local spatial-frequency moments of an image texture using Gabor elementary functions and their derivatives is presented. Also described is a technique for separating shading from texture information, which makes the shape-from-texture algorithm robust to the shading effects found in real imagery. Second, a detailed model for the projection of local spectra and spectral moments of any surface reflectance patterns (not just textures) is developed. Third, the conditions under which the projection model can be solved for the orientation of the surface at each point are explored. Unlike earlier non-feature-based, curved surface shape-from-texture approaches, the assumption that the surface texture is isotropic is not required; surface texture homogeneity can be assumed instead. The algorithms ability to operate on anisotropic and nondeterministic textures, and on both smooth- and rough-textured surfaces, is demonstrated. >

Classification of contour shapes using class segment sets

Both example-based and model-based approaches for classifying contour shapes can encounter difficulties when dealing with classes that have large nonlinear variability, especially when the variability is structural or due to articulation. This paper proposes a part-based approach to address this problem. Bayesian classification is performed within a three-level framework, which consists of models for contour segments, for classes, and for the entire database of training examples. The class model enables different parts of different exemplars of a class to contribute to the recognition of an input shape. The method is robust to occlusion and is invariant to planar rotation, translation, and scaling. Furthermore, the method is completely automated. It achieves 98% classification accuracy on a large database with many classes.

Localized measurement of image fractal dimension using gabor filters

Abstract We present a method for making accurate, optimally localized measurements of the fractal dimension of images modeled as locally fractal Brownian surfaces. Fractal Brownian surfaces are good models for the multiscale and irregular image textures arising from natural scenes and for elevation maps of terrain. To obtain accurate local values of the fractal dimension, spatio-spectrally localized measurements are necessary. Our method employs Gabor filters, which optimize the conflicting goals of spatial and spectral localization as constrained by the functional uncertainty principle. The outputs from multiple Gabor filters are fitted to a fractal power-law curve whose parameters determine the fractal dimension. The algorithm produces a local value of the fractal dimension at every point in the image. We also introduce a variational technique for enforcing a smoothness constraint on the computed fractal dimension function. This technique is implemented using an iterative relaxation algorithm. We demonstrate the method on synthetic images, real images of natural textures, and U.S. Geo-Data digital elevation maps of real terrain. We discuss the ways in which real images depart from the fractal Brownian surface model.

RETRIEVAL FROM SHAPE DATABASES USING CHANCE PROBABILITY FUNCTIONS AND FIXED CORRESPONDENCE

Similarity-based retrieval from shape databases typically employs a pairwise shape matcher and one or more indexing techniques. In this paper, we focus specifically on the design of a pairwise matcher for retrieval of 2-D shape contours. In the past, the matchers used for the one-to-many problem of shape retrieval were often designed for the problem of matching an isolated pair of shapes. This approach fails to exploit two characteristics of the one-to-many matching problem that distinguish it from the one-to-one matching problem. First, the output of shape retrieval systems tends to be dominated by matches to relatively similar shapes. In this paper, we demonstrate that by not expending computational resources on unneeded accuracy of matching, both the speed and the accuracy of retrieval can be increased. Second, the shape database is a large statistical sample of the population of shapes. We introduce a probabilistic model for exploiting that statistical knowledge to further increase retrieval accuracy. The model has several benefits: (1) It does not require class labels on the database shapes, thus supporting unlabeled retrieval. (2) It does not require feature independence. (3) It is parameter-free. (4) It has a fast runtime implementation. The probabilistic model is general and thus potentially applicable to other one-to-many matching problems.

Learning Chance Probability Functions for Shape Retrieval or Classification

Several example-based systems for shape retrieval and shape classification directly match input shapes to stored shapes, without using class membership information to perform the matching. We propose a method for improving the accuracy of this type of system. First, the system learns a set of chance probability functions (CPFs). The CPFs estimate the probabilities of obtaining a query shape with particular distances from each training example by chance. The learned CPFs are used at runtime to rapidly estimate the chance probabilities of the observed distances between the actual query shape and the database shapes. These estimated probabilities are then used as a dissimilarity measure for shape retrieval and/or nearest-neighbor classification. The CPF learning method is parameter-free. Experimental evaluation demonstrates that: (1) chance probabilities yield higher accuracy than Euclidean distances; (2) the learned CPFs support fast matching; and (3) the CPF-based system outperforms prior systems on a standard benchmark test of retrieval accuracy.

Fast correspondence-based system for shape retrieval

Several recently published shape retrieval systems have achieved high accuracy on a benchmark 1400-shape dataset. However, some of these systems have high pairwise shape matching costs due to their use of structural matching or flexible correspondence. The purpose of this paper is to demonstrate that a relatively simple shape retrieval system based on fixed point correspondences can achieve accuracy comparable with the two most accurate prior systems, at significantly higher speed. High accuracy is achieved by using a stable curve normalization procedure and example-based retrieval. High speed is achieved by three techniques: single key-point alignment, fixed correspondences, and PCA-based dimensionality reduction. The system is completely automatic and does not require that the database shapes have class labels.

Shape-from-texture by wavelet-based measurement of local spectral moments

A closed-form solution to the problem of computing 3D curved-surface shape from texture cues is presented. An expression showing the dependence of localized image spectral moments on localized surface spectral moments and on local surface orientation is derived. The local image spectra are measured with wavelets, and the expression is solved for the surface orientation at each point. Because the method uses localized spectral information, it operates at a very low level in the visual hierarchy. No extraction of texture or edge elements is required. The wavelet-based computation used is biologically plausible, easily parallelized for rapid computation, and has been shown to be the basis for effective solutions to a variety of other vision tasks. The method is demonstrated on a number of real-world examples.<<ETX>>

Fast retrieval of isolated visual shapes

Similarity-based retrieval from databases of isolated visual shapes has become an important information retrieval problem. The goal of the current work is to achieve high retrieval speed with reasonable retrieval effectiveness, and support for partial and occluded shape queries. In the proposed method, histograms of local shape parts are coded as index vectors. To increase retrieval accuracy, a rich set of parts at all scales of the shape is used; specifically, the parts are defined as connected sequences of regions in curvature scale space. To increase efficiency, structural indexing is used to compare the index vectors of the query and database shapes. In experimental evaluations, the method retrieved at least one similar shape in the top three retrieved items 99-100% of the time, depending on the database. Average retrieval times ranged from 0.7 ms on a 131-shape database to 7 ms on a 1310-shape database. The method is thus suitable for fast, approximate shape retrieval in comparison with more accurate but more costly structural matching.