Images and video captured by drones and unmanned systems can be subject to a variety of distortions such as motion blur, fisheye effect and overexposure. These distortions may arise due to internal factors such as sensor performance, or external factors such as rain, fog and wind. Graphics enhancement software can compensate for these errors and make the captured images more usable for actionable intelligence and for deliverables such as maps and orthophotos.
Due to the limitations on the flight time of many drones, imagery may be captured at as great a height as possible in order to maximise the viewing angle of the camera and thus cover a greater area in a single flight. The downside of this is that the resolution of the imagery is reduced, which in turn reduces the level of detail.
Drone image enhancement software can be used to enhance the clarity of synthetic aperture radar (SAR) imagery, optimize infrared (IR) imagery taken in low-illumination environments, and perform enhancements such as adjusting the image contrast relative to differences in temperature. It can also be used to remove noise from images, as well as other unwanted details such as shadows and haze.
Software for upscaling drone imagery can be used for a variety of applications, such as removing the effects of smoke, haze and airborne particles while tracking and monitoring fires. Graphics enhancement software for UAVs can also be used to pick out potential targets during military or search and rescue operations or to identify possible areas of concern during industrial inspection operations, such as the monitoring of pylons, bridges and pipelines.
Artificial intelligence based image enhancement software can be used to artificially upscale the imagery, with algorithms that can estimate the values of unknown pixels by using data from multiple lower-resolution images.
Video recorded by unmanned systems can be optimised to enhance presentation, transmission and display. Software for video enhancement and overlaying techniques can be used to optimise: