Visual Intelligence: iOne STKA – A Software Toolkit Platform for Developing Next Generation Drone Array Sensors
Dr. Armando Guevara, CEOAirborne drones and hand-held devices in Geographical Information Systems (GIS) have emerged as game-changing platforms with the potential to capture geospatial data in unprecedented detail, greater accuracy and more efficiently. However, the plethora of sensors/DSLR cameras available today generally underperform for many important applications including engineering analysis of critical infrastructure and the production of 3D base layers needed by next-generation augmented reality, city sustainability and utilities monitoring/ management (e.g. electrical, cell towers), as well as many other industry applications.
Visual Intelligence Inc, based out of Houston, Texas is the development bedrock of new technology centered on array-based airborne and handheld 3D data collection utilizing photogrammetry and computer vision. The technology caters to a broad range of industries, which require high quality and accurate (metric), scalable and affordable digital sensor system solutions. “We have created a software sensor toolkit architecture (a sensor operating system platform) to efficiently and economically integrate different kinds of sensors that will lead drones to dominate next-generation applications of data collection,” says Dr. Armando Guevara, CEO, Visual Intelligence.
The iOne STKA platform is a patented modular array architecture which enables end-to-end integration via the co-mounting and co-registration of similar or diverse sensor types (active or passive such as LiDAR, Radar, Thermal, IR) for 3D point clouds, DSM and depth map computation. This unique collection and functional scalability delivers the highest metric quality data products: stereo pairs with .6 b/h ratio for accurate DSMs and feature extraction; point clouds that are 10-20x+ more dense than LiDAR; positional accuracies of 3-6 inches; and, derived 3D models at 0.5 - 2 inch resolution.
The iOne STKA enables end-users and organizations to readily produce the diversity of actionable information required by their GIS applications, utilizing one software platform that supports multipurpose scalable devices from large to micro, aerial and mobile sensor systems.
Core to the technology is the sensors’ ability to generate a metric virtual image from multiple image sources, i.e. multiple economical cameras become a single more accurate and higher performance sensor; and the virtual image can generate basic 2D and 3D products.
We have created a software sensor toolkit architecture to efficiently and economically integrate different kinds of sensors that will lead drones to dominate next-generation applications of data collection
The platform converts an array of sensors into a wide field imager with metric accuracy and packs more pixels at high definition in a smaller envelope. “This camera array enables a wider field view which results in collecting more pixels, through the smallest aperture, per unit of time/area. The virtual image size being larger allows 2-3 times faster collection, yielding significant savings in collection and processing time, all while maintaining impressive to image quality and accuracy,” says Guevara.
The iOne STKA software platform is an end-to-end highly automated workflow encapsulating critical calibration, fusion and image processing, photogrammetry, and computer vision techniques. The platform architecture adapts to different sensor types enabling the best possible information product any given sensor is capable of producing. This means pixel-level positional accuracy, fusion of disparate data layers, dense point clouds, small GSD’s, and high dynamic range. The array geometry collects more pixels than a single camera and will scale as future detector arrays become increasingly dense; arrays enable richer and more accurate data quality needed by advanced applications such as 3D modeling at the engineering level; current and future aerial platforms can now create a base layer for all other 3D information. While aircraft will continue to collect large areas at the lowest cost per pixel/highest accuracy, drones enable maintenance of that information’s constant currency.
Guevara foresees Visual Intelligence becoming the leader in “intelligent metric-quality sensors—everywhere and anywhere; technology IoT connected so efficiently and economically that it will soon make the paradigm of ‘geoimaging the 3D world online’ a reality.”