Satellite Component Detection

Satellite component detection focuses on automatically identifying and classifying parts of spacecraft in images, crucial for autonomous on-orbit servicing and space debris removal. Research heavily utilizes convolutional neural networks (CNNs), often combined with techniques like 3D Gaussian splatting or traditional computer vision methods to improve accuracy and address limitations such as misclassifications. These advancements aim to enhance the reliability and safety of autonomous space operations by providing robust and accurate object detection capabilities for vision-based navigation and control systems. The development of efficient, onboard processing methods, including parameter-efficient fine-tuning, is also a key area of focus to reduce reliance on ground-based data transmission.