ABSTRACT
This chapter focuses on basic template-matching technique is investigated for the identification of circular features in imagery that is easy to implement using common GIS tools. It makes heavy use of GIS and image-processing methods and is largely experimental in nature, although a real-world case study is also offered. A central principle in aerial image interpretation is that regular geometric shapes are generally of human origin. The central idea behind optical character recognition (OCR) is pattern-matching to recognize characters in an image of a printed page. More recently, Trier, Larsen, and Solberg attempted to identify circular archaeological mounds in Ikonos satellite imagery in Norway, with good results in a wide number of images using a template-matching algorithm. The results considered here represent an attempt to automatically detect archaeological features in air or space imagery based on their regular shape characteristics.
