LiveEO is pleased to announce that InvestitionsBank Berlin (IBB) has awarded us a research and development grant under the Pro FIT programme. The project is titled “ProFIT II - Precision Satellite Analysis for Complex Environments and Target Applications” and is supported by a €277,000 grant until 2026.
Project Focus: Better Co-Registration of Satellite Imagery
The goal of this initiative is to significantly improve and automate the co-registration of stereoscopic satellite imagery. This advancement aims to minimize the need for manual horizontal and vertical offset corrections, thereby increasing the robustness and efficiency of LiveEO’s risk assessment solutions for linear infrastructure.
The Challenge: Spatial Misalignment in Stereo Imagery
When stereo image pairs are captured at different times, from varying viewpoints, or using different sensors, they often suffer from spatial misalignments. Image co-registration, a technique for minimizing these shifts, is critical for ensuring the accuracy of remote sensing analyses. Without it, displacement between images can introduce significant errors into downstream applications.
Treeline in Focus: Why Accurate Canopy Height Models Matter
LiveEO’s Treeline product monitors vegetation-related risks to power lines and railways. It uses stereo imagery to assess vegetation height along infrastructure corridors by subtracting a digital elevation model (DEM) from a digital surface model (DSM) to produce a canopy height model. This model is a key input for calculating the risk posed by surrounding vegetation.
However, if the DEM and DSM are not perfectly aligned, even small offsets can result in errors of several meters in estimated vegetation height. This is a problem that especially occurs in mountainous and forested terrain. These inaccuracies often require time-consuming manual correction, slowing project delivery.
Manual to Machine: Automating Ground Control Point Collection
Currently, co-registration relies on the manual collection of ground control points, distinct features that are visible in both images. This process is not only labor-intensive but also constrained by the precision of human clicks.
To overcome this, the new project will explore advanced machine learning methods to automate this alignment process. We will work with external partners to generate synthetic imagery, computer-generated scenes, that can increase the number of data points that can be used to train our models. The goal is to eliminate both vertical and horizontal offsets, thereby enhancing the accuracy and scalability of LiveEO’s analysis pipeline.
Accelerating Delivery: Benefits for Infrastructure Operators
Automating the co-registration process will drastically reduce the time required to deliver client projects. This will enable Treeline to support more customers, particularly large infrastructure operators, on tighter schedules, ultimately increasing LiveEO’s service capacity and impact.
Looking Ahead: Project Timeline and Strategic Impact
This project marks a major milestone in LiveEO’s mission to lead in satellite-based infrastructure monitoring. By pushing the boundaries of image processing and automation, we are laying the groundwork for faster, more scalable, and more accurate risk assessment solutions, paving the way for innovation in the infrastructure and Earth observation sectors through 2026 and beyond.
