Drone Imaging with MIPAR

Drone Imaging with MIPAR

Drone imaging is a radically disruptive technology that is faster, safer, and oftentimes more economic than manual surveying. Although camera drones and unmanned aerial vehicles (UAVs) are still largely associated with either hobbyists or the military, businesses are rapidly waking up to their enormous potential in construction, environmental screening, infrastructural inspection, manufacturing, urban planning, and more.

​

At MIPAR Image Analysis, we have partnered with several survey companies to ensure that they can reliably extract detailed measurements from aerial images. We specialize in algorithmic image analysis based on a powerful deep learning toolbox and an end-to-end consultative approach. The result is a completely custom, sophisticated analytical software package tuned for specific user requirements – even in new and emerging markets. In the last three years, drone imaging has become one of our core specialties.

Drone Imaging 101: The Basics

Briefly: Drone imaging covers any application where a camera is fitted to a lightweight unmanned aerial system (UAS) for remote photography or video capture. It is typically associated with obtaining images from height, but drones are increasingly used for low-altitude photography too. Imagine how difficult it can be for human operators to carry out process assurance and product inspections on fully-automated manufacturing lines with robotic assemblies, for instance. Drone imaging can also be useful for defect screening and product control in sensitive environments

The Matter of Drone Optics

Drone imaging systems can typically be classified by their optics, usually either visible or infrared. Aerial photography primarily concerns visible light, while thermal imaging creates pictures using radiation on the near-infrared (NIR), short-wave infrared (SWIR), and mid-wave infrared (MWIR) spectrums.

​

Hyperspectral cameras are increasingly integrated into UAV systems, but these are incredibly expensive and are yet to reach the same degree of market penetration as visual photography and thermography. There has also been a gradual uptick of industrial drones fitted with X-ray cameras – which sparked some debate over personal privacy, but the reality is that drone imaging with X-rays is currently exclusive to research environments

How to Analyze Complex Drone Imagery

Drone technology has only been available to consumers for a decade, and many business sectors are still in a first-mover phase where only one or very few companies are implementing drone imaging in any meaningful way. Despite that, UAV imaging systems have become extremely sophisticated in a short time-frame, delivering increasingly complex imagery based on practically any spectrum of interest.

You can now measure crystallinity in semiconductor defect screening using X-ray diffraction (XRD) imaging mounted on a UAV. Contaminant particles can likewise be detected and classified based on grain shape, size, and orientation. At the conventional aerial imaging level, you can carry out remote sensing and monitoring of practically any object. The problem with each of these, however, is that without sophisticated software, image analysis must be carried out on a manual basis. This sacrifices many of the autonomy benefits that make drone imaging so attractive in the first place.