Thomas K. Leung

Texture synthesis by non-parametric sampling

A non-parametric method for texture synthesis is proposed. The texture synthesis process grows a new image outward from an initial seed, one pixel at a time. A Markov random field model is assumed, and the conditional distribution of a pixel given all its neighbors synthesized so far is estimated by querying the sample image and finding all similar neighborhoods. The degree of randomness is controlled by a single perceptually intuitive parameter. The method aims at preserving as much local structure as possible and produces good results for a wide variety of synthetic and real-world textures.

Representing and Recognizing the Visual Appearance of Materials using Three-dimensional Textons

We study the recognition of surfaces made from different materials such as concrete, rug, marble, or leather on the basis of their textural appearance. Such natural textures arise from spatial variation of two surface attributes: (1) reflectance and (2) surface normal. In this paper, we provide a unified model to address both these aspects of natural texture. The main idea is to construct a vocabulary of prototype tiny surface patches with associated local geometric and photometric properties. We call these 3D textons. Examples might be ridges, grooves, spots or stripes or combinations thereof. Associated with each texton is an appearance vector, which characterizes the local irradiance distribution, represented as a set of linear Gaussian derivative filter outputs, under different lighting and viewing conditions.Given a large collection of images of different materials, a clustering approach is used to acquire a small (on the order of 100) 3D texton vocabulary. Given a few (1 to 4) images of any material, it can be characterized using these textons. We demonstrate the application of this representation for recognition of the material viewed under novel lighting and viewing conditions. We also illustrate how the 3D texton model can be used to predict the appearance of materials under novel conditions.

Contour and Texture Analysis for Image Segmentation

This paper provides an algorithm for partitioning grayscale images into disjoint regions of coherent brightness and texture. Natural images contain both textured and untextured regions, so the cues of contour and texture differences are exploited simultaneously. Contours are treated in the intervening contour framework, while texture is analyzed using textons. Each of these cues has a domain of applicability, so to facilitate cue combination we introduce a gating operator based on the texturedness of the neighborhood at a pixel. Having obtained a local measure of how likely two nearby pixels are to belong to the same region, we use the spectral graph theoretic framework of normalized cuts to find partitions of the image into regions of coherent texture and brightness. Experimental results on a wide range of images are shown.

Textons, contours and regions: cue integration in image segmentation

The paper makes two contributions: it provides (1) an operational definition of textons, the putative elementary units of texture perception, and (2) an algorithm for partitioning the image into disjoint regions of coherent brightness and texture, where boundaries of regions are defined by peaks in contour orientation energy and differences in texton densities across the contour. B. Julesz (1981) introduced the term texton, analogous to a phoneme in speech recognition, but did not provide an operational definition for gray-level images. We re-invent textons as frequently co-occurring combinations of oriented linear filter outputs. These can be learned using a K-means approach. By mapping each pixel to its nearest texton, the image can be analyzed into texton channels, each of which is a point set where discrete techniques such as Voronoi diagrams become applicable. Local histograms of texton frequencies can be used with a /spl chi//sup 2/ test for significant differences to find texture boundaries. Natural images contain both textured and untextured regions, so we combine this cue with that of the presence of peaks of contour energy derived from outputs of odd- and even-symmetric oriented Gaussian derivative filters. Each of these cues has a domain of applicability, so to facilitate cue combination we introduce a gating operator based on a statistical test for isotropy of Delaunay neighbors. Having obtained a local measure of how likely two nearby pixels are to belong to the same region, we use the spectral graph theoretic framework of normalized cuts to find partitions of the image into regions of coherent texture and brightness. Experimental results on a wide range of images are shown.

Recognizing surfaces using three-dimensional textons

We study the recognition of surfaces made from different materials such as concrete, rug, marble or leather on the basis of their textural appearance. Such natural textures arise from spatial variation of two surface attributes: (1) reflectance and (2) surface normal. In this paper, we provide a unified model to address both these aspects of natural texture. The main idea is to construct a vocabulary of prototype tiny surface patches with associated local geometric and photometric properties. We call these 3D textons. Examples might be ridges, grooves, spots or stripes or combinations thereof Associated with each texton is an appearance vector, which characterizes the local irradiance distribution, represented as a set of linear Gaussian derivative filter outputs, under different lighting and viewing conditions. Given a large collection of images of different materials, a clustering approach is used to acquire a small (on the order of 100) 3D texton vocabulary. Given a few (1 to 4) images of any material, it can be characterized using these textons. We demonstrate the application of this representation for recognition of the material viewed under novel lighting and viewing conditions.

Finding Pictures of Objects in Large Collections of Images

Retrieving images from very large collections, using image content as a key, is becoming an important problem. Users prefer to ask for pictures using notions of content that are strongly oriented to the presence of abstractly defined objects. Computer programs that implement these queries automatically are desirable, but are hard to build because conventional object recognition techniques from computer vision cannot recognize very general objects in very general contexts. This paper describes our approach to object recognition, which is structured around a sequence of increasingly specialized grouping activities that assemble coherent regions of image that can be shown to satisfy increasingly stringent constraints. The constraints that are satisfied provide a form of object classification in quite general contexts. This view of recognition is distinguished by: far richer involvement of early visual primitives, including color and texture; hierarchical grouping and learning strategies in the classification process; the ability to deal with rather general objects in uncontrolled configurations and contexts. We illustrate these properties with four case-studies: one demonstrating the use of color and texture descriptors; one showing how trees can be described by fusing texture and geometric properties; one learning scenery concepts using grouped features; and one showing how this view of recognition yields a program that can tell, quite accurately, whether a picture contains naked people or not.

Contour Continuity in Region Based Image Segmentation

Region-based image segmentation techniques make use of similarity in intensity, color and texture to determine the partitioning of an image. The powerful cue of contour continuity is not exploited at all. In this paper, we provide a way of incorporating curvilinear grouping into region-based image segmentation. Soft contour information is obtained through orientation energy. Weak contrast gaps and subjective contours are completed by contour propagation. The normalized cut approach proposed by Shi and Malik is used for the segmentation. Results on a large variety of images are shown.

Detecting, localizing and grouping repeated scene elements from an image

This paper presents an algorithm for detecting, localizing and grouping instances of repeated scene elements. The grouping is represented by a graph where nodes correspond to individual elements and arcs join spatially neighboring elements. Associated with each arc is an affine map that best transforms the image patch at one location to the other. The approach we propose consists of 4 steps: (1) detecting “interesting” elements in the image; (2) matching elements with their neighbors and estimating the affine transform between them; (3) growing the element to form a more distinctive unit; and (4) grouping the elements. The idea is analogous to tracking in dynamic imagery. In our context, we “track” an element to spatially neighboring locations in one image, while in temporal tracking, one would perform the search in neighboring image frames.

Context-aided human recognition – clustering

Context information other than faces, such as clothes, picture-taken-time and some logical constraints, can provide rich cues for recognizing people. This aim of this work is to automatically cluster pictures according to persons identity by exploiting as much context information as possible in addition to faces. Toward that end, a clothes recognition algorithm is first developed, which is effective for different types of clothes (smooth or highly textured). Clothes recognition results are integrated with face recognition to provide similarity measurements for clustering. Picture-taken-time is used when combining faces and clothes, and the cases of faces or clothes missing are handled in a principle way. A spectral clustering algorithm which can enforce hard constraints (positive and negative) is presented to incorporate logic-based cues (e.g. two persons in one picture must be different individuals) and user feedback. Experiments on real consumer photos show the effectiveness of the algorithm.

Image and video segmentation: the normalized cut framework

We propose a segmentation system based on the normalized cut framework proposed by Shi and Malik (see Proc. IEEE Conf. Computer Vision and Pattern Recognition, San Juan, Puerto Rico, p.731-7, 1997). The goal is to partition the image from a big picture point of view. Perceptually significant groups are detected first while small variations and details are treated later. Different image features-intensity, color, texture, contour continuity, motion and stereo disparity are treated in one uniform framework.