New software tool aims to democratize access to space camera development

A powerful software tool capable of accurately modeling how cameras capture light could help democratize the development of new imaging systems for use in space.

Dr. George Brydon of the University of Glasgow developed the open-source tool, called SIMply, which is available as a free download on the online code repository GitHub.

SIMply aims to provide scientists and engineers with easy access to a level of sophisticated image simulation that is traditionally available only to large corporations or other highly funded research and development organizations with access to specialized tools.

The tool is capable of virtually modeling the interactions between light, surfaces, lenses and sensors to demonstrate the capabilities of a wide variety of camera systems, helping researchers to validate the performance of prototype designs before moving to hardware development.

It uses sophisticated ray-tracing and physical modeling techniques to render scenes accurately based on data provided by the user. It can simulate the noise created by camera sensors, and mimic arbitrary lens projections, from pinhole and fisheye to panoramic scanning cameras.

In a new paper published in the journal Space: Science & Technology, Dr. Brydon shows how he validated SIMply’s ability to accurately model photographs from space by comparing the simulations it created to images taken by spacecraft of real extraterrestrial surfaces.

He used high-resolution 3D data collected by previous missions to the moon, the asteroid Itokawa, and the comet 67P/Churyumov-Gerasimenko, to create detailed simulations of the light reflected from their surfaces and captured by the spacecraft’s onboard cameras.

In the paper, Dr. Brydon used analysis of the simulated images’ physical brightnesses and image features to demonstrate that SIMply is capable of producing simulated images which replicate the real photographs with a high physical accuracy when the input data is of sufficient quality.

When provided with highly detailed input data, the software can produce images which are visually indistinguishable from real images. The paper demonstrates this with a simulation of an image of the moon taken by the Rosetta mission in 2007 which recreated the reflectivity, details of the moon’s topography and camera sensor response with a high level of accuracy.

Dr. Brydon, a collaborative research fellow in the University of Glasgow’s School of Physics & Astronomy, said, “The early stages of SIMply’s development began during my Ph.D., when I realized that there were no freely available tools to help me model novel spacecraft cameras.

“Large companies working on camera development have access to a wide range of powerful tools because they have the resources to make their own software, but those tools remain private and inaccessible to other researchers.

“My objective in fully developing SIMply using the Python programming language was to make a tool that’s widely accessible to researchers in the planetary science and spacecraft design field. It’s available for free use to everyone from university students all the way to engineers at SMEs with great ideas who otherwise don’t have access to the resources needed to fully model their potential in software.

“The paper shows clearly how useful it could be for space applications, where the cost of developing hardware and software is so high that there’s a real advantage to using fast, free image simulation to support design and testing. I hope that making it freely available and open-source will enable others to use it for their research and development projects, and facilitate new imaging capabilities.”

While the paper focuses on SIMply’s ability to model planetary surfaces, the tool’s sophisticated modeling could make it useful for a wide range of applications in computer vision development. Dr. Brydon believes that the tool could find use in the development of cameras for use in self-driving cars, or for ball-tracking technologies for sports such as football, golf or tennis.

He is currently working to add additional camera models and enhanced capabilities to SIMply, as well as improving the supporting documentation to help new users adopt the technology for their own projects.

SIMply is currently supporting collaborative research between Dr. Brydon and Dr. Divya Persaud of the University of Glasgow’s School of Geographical & Earth Sciences, combining planetary geology with camera system modeling to develop new imaging techniques for planetary surface science.