Joseph L. Mundy and Andrew Zisserman,
Towards a New Framework for Vision,
In Joseph Mundy and Andrew Zisserman eds., Geometric Invariance in Computer Vision,
Cambridge, MA, USA, Series: Artificial intelligence, pp. 1-39, The MIT Press, 1992.
Joseph L. Mundy and Andrew Zisserman,
Towards a New Framework for Vision,
In Joseph Mundy and Andrew Zisserman eds., Geometric Invariance in Computer Vision,
Cambridge, MA, USA, Series: Artificial intelligence, pp. 1-39, The MIT Press, 1992.The importance of invariance to machine vision has been recognized since the origin of the field in the 1960s. Initially, the emphasis was on invariance to photometric properties and this provided the impetus for the development of edge detectors. In this book we are concerned with geometric invariance. Invariants are properties of geometric configurations which remain unchanged under an appropriate class of transformations. In many case the only properties of a configuration that are of any interest are those that are invariant under a particular set of transformations. The straight line is the basic feature used in vision systems precisely because it is the simplest projective invariant - an imaged straight line is straight. In this overview we outline the applications of geometric invariants in machine vision and provide a brief review of the principles of invariance. The remainder of the introduction examines these issues in more detail and provides a guide to current and future applications of geometric invariance in machine vision.