Simple algorithm for detection of elliptical objects in remotely sensed images for UAV applications

摘要

Elliptical object recognition in remotely sensed images has been an active research problem in the field of computer vision. In this research work an efficient algorithm for elliptical object detection from remotely sensed image (aerial and satellite images) is proposed which has very important application in Unmanned Aerial Vehicle (UAV) target and path detection applications. A comparative analysis between our proposed and already exiting algorithms is made. It is observed that the proposed algorithm-technique is efficient, simple, easy to implement and has the ability to recognize elliptical objects, deformed objects with nearly elliptical-circular shape. The proposed object identification and extraction algorithm is based on Line Following Technique (LFT), linear correlation based analysis of Euclidian distance trend of elliptical, circular objects recognition, identification of elliptical, circular, close to circular objects is made. It is concluded that the proposed algorithm is easy to implement with comparable level of accuracy, speed and can also be use for detection in a low resolution image with accuracy.