David H. Cooper

Active shape models—their training and application

!, Model-based vision is firmly established as a robust approach to recognizing and locating known rigid objects in the presence of noise, clutter, and occlusion. It is more problematic to apply modelbased methods to images of objects whose appearance can vary, though a number of approaches based on the use of flexible templates have been proposed. The problem with existing methods is that they sacrifice model specificity in order to accommodate variability, thereby compromising robustness during image interpretation. We argue that a model should only be able to deform in ways characteristic of the class of objects it represents. We describe a method for building models by learning patterns of variability from a training set of correctly annotated images. These models can be used for image search in an iterative refinement algorithm analogous to that employed by Active Contour Models (Snakes). The key difference is that our Active Shape Models can only deform to fit the data in ways consistent with the training set. We show several practical examples where we have built such models and used them to locate partially occluded objects in noisy, cluttered images. Q 199s A&& prrss, IN.

Training Models of Shape from Sets of Examples

A method for building flexible shape models is presented in which a shape is represented by a set of labelled points. The technique determines the statistics of the points over a collection of example shapes. The mean positions of the points give an average shape and a number of modes of variation are determined describing the main ways in which the example shapes tend to deform from the average. In this way allowed variation in shape can be included in the model. The method produces a compact flexible ‘Point Distribution Model’ with a small number of linearly independent parameters, which can be used during image search. We demonstrate the application of the Point Distribution Model in describing two classes of shapes.

Trainable method of parametric shape description

We have developed a trainable method of shape representation which can automatically capture the invariant properties of a class of shapes and provide a compact parametric description of variability. We have applied the method to a family of flexible ribbons (worms) and to heart shapes in echocardiograms. We show that in both cases a natural parameterisation of shape results.

Building and using flexible models incorporating grey-level information

The authors describe a technique for building compact models of the shape and appearance of flexible objects seen in 2-D images. The models are derived from the statistics of sets of labeled images of example objects. Each model consists of a flexible shape template, describing how important points of the object can vary, and a statistical model of the expected grey levels in regions around each model point. Such models have proved useful in a wide variety of applications. A description is given on how the models can be used in local image search, and examples of their application are included.<<ETX>>

The use of geometric and grey-level models for industrial inspection

Abstract A methodology is described for reliably locating pre-defined features in constrained images. The method uses a geometric model incorporating knowledge about the images involved, and a grey-level model capable of identifying specific features in a grey level profile. Results from a real application are presented.

A compact set of image processing primitives and their role in a successful application program

Abstract We describe an integrated set of data structures and image processing primitives which enhance not only the run time efficiency of the application program, but also the programmability of the solution. This is illustrated by a program for clinical chromosome analysis.

An object location strategy using shape and grey-level models

Abstract Any model-based image interpretation system must be capable of describing objects, whose appearance in real images can vary widely, in sufficient detail to ensure that robust location of objects is possible. The system must cope with circumstances where data is incomplete, for example when touching and occlusion occur. It is argued that to achieve this it is necessary to describe grey-level properties as well as geometric ones, and their expected variations. This paper proposes an object description which combines an explicit shape model with models of expected grey-level boundary appearance together with a mechanism for evaluating image data for correspondences to the model. The results of applying the method to locating the boundaries of overlapping and touching objects in microscope images of metaphase chromosomes and manmade mineral fibres are presented.

Image search using trained flexible shape models.

This paper describes a technique for building compact models of the shape and appearance of flexible objects seen in two-dimensional images. The models are derived from the statistics of sets of images of example objects with ‘landmark’ points labelled on each object. Each model consists of a flexible shape template, describing how the landmark points can vary, and a statistical model of the expected grey levels in regions around each point. Such models have proved useful in a wide variety of applications. We describe how the models can be used in local image search and give examples of their application.

Shape verification using belief updating

This paper addresses the problem of2D shape representation and its application to object verification. We show how knowledge of shape can be integrated in a principled manner with low-level evidence such as an estimate of object position and an edge strength map. We begin by considering the role of shape in image interpretation and discuss the criteria which should be applied in assessing representations of shape. We propose new criteria, particularly as regards the ability to model variability, and describe a Chord Length Distribution (CLD) representation of shape which possesses many desirable properties. We show how the CLD representation can be used in an iterative belief-updating scheme for object location and verification. We give experimental results which demonstrate the feasibility of the method and discuss future developments.

Object cues for model-based image interpretation

This paper is concerned with generating object cues from grey-level images for use in model-based image interpretation. We describe the idea of local grey-level symmetry and illustrate how points in the grey-level image with this property form local axes of symmetry. These axes together with appropriate scale information form the object cues. The degree of local symmetry in the grey-level image is made explicit by introducing a Centre of Gravity filter. The local axes of symmetry are shown to appear in the centre of gravity image as troughs and a method for labelling the troughs is described. We give results for real images and make an objective comparison with other published methods.