Review Active Shape Models - Part II: Image Search and Classification
Rafeef Abugharbieh, Ghassan Hamarneh, Tomas Gustavsson · 1998
In this paper we review and investigate a method capable of modeling the different appearance of objects in images that is due to natural shape variations, varying lighting conditions, 3D pose and others. Objects are represented by well-defined landmark points and shape variations are modeled using a principal component analysis. Also, gray level variations are being modeled. The first part of the paper describes the shape and gray scale modeling in some detail. The second part describes an iterative algorithm which deforms an initial model to fit data in ways that are consistent with shape variations found in previously acquired training data. An application to image classification is outlined. 1. Introduction Given a training set of accurately aligned shapes of an object class, we can model these shapes by xxbP=+ where x is the mean shape, P is the matrix of the t first principal components, and b is a vector of weights. We refer to this model as a Point Distribution Model (PDM). ...