REAL-TIME DATA PROCESSING IN ORBIT


Unibap's LOOM enables satellites to autonomously process hyperspectral data directly in orbit, delivering actionable insights in real time. This low-latency capability enhances situational awareness and supports time-critical decision making processes.

CUSTOMIZABLE AND EFFICIENT DATA DELIVERY


Unibap LOOM can easily be adapted to handle different number of bands and image sizes. Its sensor-agnostic design ensures compatibility with different hyperspectral sensors, abstracting away hardware complexity while maintaining a uniform, high-quality output. This allows seamless integration into existing satellite constellations and AI workflows.

ENHANCED IMAGE QUALITY AND AI READINESS


Unibap LOOM abstracts away the sensor, creating the same quality imagery regardless which series, model or brand of sensor that is connected to its inlet. This consistency enables AI models trained on one spacecraft to be used with the same performance on another without costly re-training and optimization.

Trusted by

Want to know more?

KEY TECHNICAL SPECIFICATIONS


Input Methods: Directly from sensor or from disk
Input Data Format: Packetized RAW sensor data
Output Format: Corrected GEO-TIFF images
Max Number of Bands: Up to 32 bands (more bands optional)
Pixel Width: 4096 pixels (wider arrays optional)
Processing Mode: Multi-threaded, runs as a single binary
Georeferencing: UTM coordinate-based affine transformation
Installation Methods: .deb packages for Linux-based systems

IMAGE CORRECTION CAPABILITIES


Correction type
Band Co-Registration: Aligns spectral bands to remove distortions
De-Noising: Reduces sensor noise via statistical correction
Radiance to Reflectance: Converts raw radiance values for accurate data

DEPLOYMENT MODES


Standalone Binary: Direct processing on raw sensor data
As a Service: Integrates with Unibap SCOS (only for sensor input method)

Downloads