False Alarm Reduction in Wavelength–Resolution SAR Change Detection Schemes by Using a Convolutional Neural Network

In this letter, we propose a method to reduce the number of false alarms in wavelength–resolution synthetic aperture radar (SAR) change detection scheme by using convolutional neural network (CNN). The is performed two steps: analysis and object classification. A simple technique for SAR implemented extract potential targets from image interest. CNN then used classifying map detections as either target or nontarget, further reducing alarm rate (FAR). tested CARABAS-II data set, where only three over testing area 96 km 2 are reported while still sustaining probability above 96%. We also show that can FAR even when flight heading system measurement campaign differs up 100° between images training test.