Authors : Kayvan Najarian , Robert DeMott II and Jeremy Cooper
Abstract: The use of multiple shape recognition techniques for target recognition and classification are explored and compared. An overview into the methods used to isolate the targets and remove background clutter is also included. The primary method used for shape detection was a version of geometric hashing, used in combination with a specialized corner detection method and a circle detection method. The corner detection method described obtained better results at finding corners in the black and white outlined images than several other compared methods. The geometric hashing algorithm implemented with a weighted-neighbourhood, performed better than the alternative methods compared.
Kayvan Najarian , Robert DeMott II and Jeremy Cooper , 2008. Automatic Target Recognition and Classification in Aerial Photography. International Journal of Signal System Control and Engineering Application, 1: 1-14.