Enhancing military capabilities with technology
With AI transforming the defense landscape around the world, nations are heavily investing in AI powered technologies to enhance their military capabilities. Recognising its potential in various aspects of warfare, from cyber defense systems to autonomous systems, European nations have significantly increased investments in AI defense technologies. A key driver for this shift is the nation’s desire to reduce reliance on external powers and build technological sovereignty.
As a new age defense AI company developing AI-based solutions to protect democracies, the client required accurate annotations on SAR images to support their Artificial intelligence project. This included accurately annotating visible features on synthetic-aperture radar (SAR) images and manually annotating the features using RGB images as reference. Categorising annotations into predetermined classes, such as terrain, infrastructure, and vehicles, was crucial. Additionally, it was important to maintain a high level of accuracy in both spatial placement and classification.
RMSI developed an accurate solution for annotating features on SAR (Synthetic Aperture Radar) images within the project. Our team utilized the client’s web-based AI/ML platform for image annotation, collaboration, and data management. The workflow involved training the team to understand project specifications and quality standards. During the annotation phase, features on SAR images were manually marked, guided by reference RGB images. These annotations were grouped into specific classes and subdivisions. Rigorous quality checks were performed to ensure the annotations achieved a high accuracy level of 95%. Finally, the annotated SAR images were delivered to the client.
As per the project requirements, RMSI delivered accurate annotations that met the high standards necessary for AI and Ministry of Defence applications. The project ensured comprehensive coverage of all features in the SAR images and achieved precise classification accuracy according to predetermined classes. The annotated data obtained could be seamlessly integrated into the client’s AI project, enhancing their capabilities and decision-making processes.
