Compress-Align-Detect: Onboard Change Detection From Unregistered Images

Change detection (CD) from satellite images typically incurs a delay ranging from several hours up to days because of latency in downlinking the acquired images and generating orthorectified image products at the ground stations; this may preclude real or near real-time applications. To overcome this limitation, we propose shifting the entire CD workflow onboard satellites. This requires simultaneously solving challenges in data storage, image registration, and CD with a strict complexity constraint. In this article, we present a novel and efficient framework for onboard CD that addresses the aforementioned challenges in an end-to-end fashion with a deep neural network composed of three interlinked submodules: 1) image compression, tailored to minimize onboard data storage resources; 2) lightweight coregistration of nonorthorectified multitemporal image pairs; and 3) a novel temporally invariant and computationally efficient CD model. This is the first approach in the literature combining all these tasks in a single end-to-end framework under the constraints dictated by onboard processing. Experimental results compare each submodule with the current state-of-the-art and evaluate the performance of the overall integrated system in a realistic setting on low-power hardware. Compelling CD results are obtained in terms of F1 -score as a function of compression rate, sustaining a throughput of 0.7 Mpixel/s on a 15-W accelerator.