![[2602.20709] Onboard-Targeted Segmentation of Straylight in Space Camera Sensors](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.20709] Onboard-Targeted Segmentation of Straylight in Space Camera Sensors
Summary
This paper presents an AI-driven methodology for segmenting straylight effects in space camera sensors, enhancing image analysis in resource-constrained environments.
Why It Matters
The research addresses a critical challenge in space imaging by improving the accuracy of fault detection in camera systems. This is essential for missions where image quality is paramount, and it leverages AI to optimize performance despite hardware limitations.
Key Takeaways
- Introduces a novel AI methodology for segmenting straylight in space cameras.
- Utilizes a pre-trained DeepLabV3 model to enhance generalization across diverse flare textures.
- Develops custom metrics to evaluate model performance within spacecraft systems.
Computer Science > Computer Vision and Pattern Recognition arXiv:2602.20709 (cs) [Submitted on 24 Feb 2026] Title:Onboard-Targeted Segmentation of Straylight in Space Camera Sensors Authors:Riccardo Gallon, Fabian Schiemenz, Alessandra Menicucci, Eberhard Gill View a PDF of the paper titled Onboard-Targeted Segmentation of Straylight in Space Camera Sensors, by Riccardo Gallon and 3 other authors View PDF HTML (experimental) Abstract:This study details an artificial intelligence (AI)-based methodology for the semantic segmentation of space camera faults. Specifically, we address the segmentation of straylight effects induced by solar presence around the camera's Field of View (FoV). Anomalous images are sourced from our published dataset. Our approach emphasizes generalization across diverse flare textures, leveraging pre-training on a public dataset (Flare7k++) including flares in various non-space contexts to mitigate the scarcity of realistic space-specific data. A DeepLabV3 model with MobileNetV3 backbone performs the segmentation task. The model design targets deployment in spacecraft resource-constrained hardware. Finally, based on a proposed interface between our model and the onboard navigation pipeline, we develop custom metrics to assess the model's performance in the system-level context. Comments: Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) Cite as: arXiv:2602.20709 [cs.CV] (or arXiv:2602.20709v1 [cs.CV] for this...