September 12, 2006
Written by Leonard David
Don Mitchell of Redmond, Washington is a retired researcher from Bell Labs and Microsoft Research. He has matched his computer science and image processing skills with a passion to study old Soviet spacecraft data.
One result of this high-tech harmony of interests has been a relook at imagery relayed via twin Venera-13 and 14 probes that landed on Venus in 1982.
Processing of the Venera lander imagery was done in several stages, Mitchell explained.
The biggest task was first taking multiple transmissions—live and from tape—then merging them to produce one very clean master copy of each of four spacecraft cameras. Venera-13 and 14 each had two cameras.
A new camera calibration function was calculated. That conversion teased out a lot of the very dark and very light regions caught by Venera cameras, Mitchell said, detail not brought out in the original Russian photo reduction work.
Surprisingly, in the case of Venera-13, distant hills hazily seen were clearly revealed.
To produce the images, Mitchell mixed in in special purpose source code, resampling, and other image wizardry, along with knowledge about dimensions of the Venera lander and location of its camera lens. He was able to preserve the original sharpness of the Venera surface pictures. A series of projected views were combined in Photoshop. A few missing spots at the horizon were filled in by duplicated or reversing other portions of the image.