AUTOMATED CO-REGISTRATION OF MULTITEMPORAL AIRBORNE FRAME IMAGES FOR NEAR REAL-TIME CHANGE DETECTION

摘要

Advances in aerial platforms, imaging sensors, image processing/computing, geo-positioning systems, and wireless communications make near real-time change detection more practical and cost effective. In Coulter et al. (2003) and Stow et al. (2003), we documented a method for collecting and precisely co-registering multitemporal airborne frame imagery that minimizes terrain and image features distortion differences between images, and enables precise relative alignment using simple techniques. This patent pending method is referred to as frame center matching. Here we describe specific procedures for automating the co-registration of frame center matched images. This type of rapid and automated image co-registration may support real-time or near real-time change detection, where newly acquired and previously acquired frame center matched images are aligned and compared on-board as an aircraft flies. The steps for automation include pairing multitemporal image sets, utilizing airborne Global Positioning System (GPS) and inertial measurement unit (IMU) data for initial alignment of images, automated control point collection, evaluation of control point pairs, image transformation, and assessment of image co-registration quality. This work is developed by the National Center for Border Security and Immigration: A Department of Homeland Security (DHS) Science and Technology (S&T) Center of Excellence. This work is also developed by the Infrastructure Protection and Disaster Management Division of DHS S&T.