Artic Small-Unmanned Aerial System Automatic Ground Control Point Processing for Terrain Modeling

esc Aerospace and Photon-X have teamed to provide a unique, low SWaP-C vision system with incorporated artificial intelligence and machine learning (AI&ML) capability that enables data collection and orthomosaic creation in featureless environments.At the core of this capability is Photon-X’s patented 3D shape-based AI&ML Spatial Phase Imaging (SPI)technology. SPI provides an unmatched AI&ML capability to identify, classify, and track small “objects.” From aerial and ground targets at distances greater than 50 KM, to closeup inspection of small imperfections in surfaces, to ridges on a fingerprint, this innovative and unique sensor system is a perfect match of this topic application.The SPI software executes directly and measures the direct surface and its orientation. Each pixel on the array becomes a multi-vector tensor with X,Y,Z, Index, specular/diffuse, etc. which creates a rapid and efficient surface algorithm. What makes SPI different?SPI is a patented technology to extract a full 3D surface from a sensor with angular information at each pixel. Each super pixel determines the pixel orientation and then calculates the pixel-to-pixel surface flows. Both LiDAR and stereo optical vision sensors use a method to extract a “point-cloud” either based on reflected light or correspondence points. This makes these technologies less efficient and less accurate. Unlike a more traditional point-cloud visual representation, SPI goes directly to the surface and pixel relationships to one another. We have then developed algorithms to measure real-time “surface flow.” In doing so, we can easily identify any changes in pixel relationships/slopes of individual surfaces and edges on a surface/object. We can operate with real-time normals and corrected for lighting and specular diffuse surfaces such as snow fields and other “featureless” terrains.