Object Detection Models Sensitivity & Robustness to Satellite-based Adversarial Attacks

The use of object detection algorithms for the analysis of satellite imagery is increasing in various fields, including environment and defense, as they enable the automatic detection, recognition and localization of targets. Satellite images often exhibit significant variations, including differences in resolution and noise levels between different satellites. Additional distortions can be caused by factors such as the position of the satellite and the specific area being scanned, resulting in changes in tangential distortion, brightness and saturation. Depending on the severity, these variations can affect the visual clarity of objects in the images and thus impair the effectiveness of object detection algorithms. This study therefore investigates the effects of such fluctuations on the performance of 3 categories of object recognition algorithms - YOLO, FASTER-RCNN and RT-DETR - by applying the principle of adversarial attacks to the inference phase of the algorithms. This experiment makes it possible to uncover the weaknesses of the algorithms facing different degrees of 5 types of satellite images noises. The case study presented is based on the automatic detection of three types of oil and gas infrastructure: compressor, tank and well in the Permian Basin (USA).