Aerial sensor technology: Major advances in efficiency and quality

Alexander Wiechert CEO, Vexcel Imaging GmbH

The aerial mapping industry experienced a notable transformation as digital sensors replaced film cameras as the preferred imaging tool, and end-to-end digital workflows became the norm. The impact of this paradigm shift continues to be felt throughout the surveying and mapping profession as more data at higher resolution and better accuracy is available, and highly automated post-processing allows faster turn-around and delivery.

Aerial mapping over the years

The motivation that drives continuous improvement of aerial imaging processes is not much different today than it was 100 years ago. How can we collect more data at higher resolution in less time? The evolution of aerial imaging technology has been influenced by a demand for efficiency and quality, and the resulting innovations have greatly improved access to high-resolution mapping data around the world.

Precursors to the modern aerial sensor, such as film cameras strapped to pigeons and suspended from kites, demonstrate the strong desire people had to document the Earth from the air long before it was feasible on a large scale. As photographic technology improved in the late 1800’s and airplanes successfully flew in the early 1900’s, the opportunities to create maps from aerial photos grew and demand increased.

Film technology evolved from panchromatic to color infrared, and as computers became available, scanning photographs to create digital files supported valuable new methods of analysis and photogrammetry. Since the introduction of the first digital aerial cameras in 2000, a constant stream of innovations has improved flying efficiency, geometric accuracy and image quality.

Key hardware enhancements

Updated versions of cameras are reinventing the mapping industry through a combination of interconnected factors, such as footprint size, GNSS/ IMU positioning, and solid-state data storage. There isn’t one single innovation that is responsible for the significant gains in aerial imaging performance; there is a cumulative positive effect.

A primary influence on flying efficiency is the size of the camera footprint. In the early days of digital cameras, the largest footprint on a digital aerial camera was 100 Megapixels, as compared to today with the UltraCam Eagle Mark 3 offering an ultra-large footprint of 450 Megapixels.

This increase is made possible by a smaller pixel design – 4 μm on the UltraCam Eagle compared to 7.2 μm on the UltraCam D camera released in 2003 – which allows more pixels, thus achieving a higher image resolution (up to 2 cm GSD). Also, as electronics on the sensors become faster, the pixels can be read more quickly. The faster frame rate allows the plane to fly at higher speeds and cover more ground, increasing productivity.

New applications that require different types of cameras, such as oblique, wide-area and hyperspectral, are also driving hardware development, and exchangeable lens kits with different focal lengths provide flexibility to vary altitudes depending on the application. Better GNSS/IMU technology is improving positioning accuracy, while improved computer components are becoming smaller, which results in lighter cameras with a smaller form factor and new options for combining several sensors in one airplane.

Onboard storage also impacts productivity. The larger volume of data being collected with more efficient sensors (up to 10 TB in a day) creates a need for lighter solid-state storage units with higher capacity that can be swapped onboard and then shipped overnight to a processing facility.

Processing improvements

Cameras with large footprints collect an incredible amount of data – and data needs to be ingested, stored, and processed. Larger file sizes and overall increased volume forced the development of better processing software that could deal with images more efficiently. Today highly automated workflows and easyto- use graphical user interfaces expedite the production of deliverables. For example, the UltraMap visualization engine Dragonfly is based on Microsoft Seadragon technology and improves the handling of UltraCam image data by using tiled images. Image pyramids and graphics card acceleration allows fast access to multi-resolution image data.

Aerial imaging in the future

With further growth of digital technology and computer processing, the entire aerial imaging workflow – from data collection to delivery – will continue to improve in efficiency and quality. As innovations are implemented, airplanes will fly higher and faster, the time required for largearea collections will be reduced, and processing will become more automated.

The demand for high-resolution imagery will increase thanks to the ongoing trend toward applications requiring high-definition maps and 3D models. Specifically, cities recognize the need for error-free detailed maps for new applications, such as autonomous navigation, and they are beginning to drive efforts to collect up-to-date accurate data with combinations of sensors, e.g., LiDAR, oblique, nadir, multispectral.

Other inventions could significantly change how we create maps altogether. Unmanned solar-powered airplanes flying 24/7 might collect data continuously during the day and transmit data at night. Persistent monitoring by fleets of small satellites may enhance the survey-grade mapping provided by aerial sensors. The wide range of possibilities is very exciting!