Space Probes

Viking 1 & 2- 1976/80 - "Project Viking"

Viking landers processed images

Image processing

Color images generated from Red, Green and Blue filter raw images (available from the Planetary Image Atlas ). Colors were calibrated by mixing the channels so that the spectral response matches that of the human eye.

See next table for details:

Viking lander Human vision
Red 700 650
Green 550 510
Blue 470 440

Viking's Red filter is shifted towards the IR part of the spectrum and this affects the final images by introducing a purple cast to the image, specially visible on the blue target and on the lander itself. The Green is close to the human eye, but still a little off towards the yellow. The Blue filter is too "short" and covers the cyan part of the spectrum. This shifts are reflected on the "Filter" text colors on the "Viking Landers" table.

This can be corrected by mixing (or averaging) the channels using proper ratios, thus obtaining a human vision like spectral response:

0.7 x RED + 0.3 x GREEN = 0.7 x 700 + 0.3 x 550 = 655
0.5 x GREEN + 0.5 x BLUE = 0.5 x 550 + 0.5 x 470 = 510
BLUE = 470 = 470


Unfortunately the blue channel is still a little off but this cannot be corrected, because we lack a "violet" or UV channel with a shorter wavelength range.

Besides addressing the issues of simple color channel balancing, light levels must also be addressed. The Viking cameras had several gain settings that allowed them to operate with different ambient light intensity. I took this into account in the processing and images are displayed with a uniform brightness level. This allows the comparison between images taken at different times of the day or to see the gradual sky and surface brightening at sunrise.

The general surface and sky colors were matched to excellent and very accurately calibrated Mer rover images by Daniel Crotty. This are, to my knowledge, the best "Mars True Color" images available. The Viking images are already very close to this and don't require much processing to adopt this corrected color. Only some small histogram balancing for white balance is enough.

Some problems to occur when the original data values are clipped or overexposed. This happens sometimes with the sky, and can mislead the viewer with a dull monotone very bright orange/brown sky. 

It's also possible to create a LRGB composition using the high resolution channel "stripe" and the wide angle color images (using Ted Stryk's super resolution images).. Data registration is very good and individual rock color variations are apparent. The only limitation to this technique is the availability of matching light conditions between the images.

The results of all of this procedures can be seen bellow. Not only do they agree with the MER rover images but the Viking color targets resemble those of Spirit and Opportunity.The blue calibration square is still a little magenta, but with no proper blue filter on camera that can be expected. Also, the blue paint looks a little pink even on earth, from the photos I've seen of the landers on test or in museum displays.

The sky and rock colors look consistent and the Viking 1 site resemble the Pathfinder location. This is not surprising, as they are relatively close on Mars and on the same type of terrain (broadly speaking). 

Finally, I've put together some polar projection 360º views of the landing sites from different Nasa imagery. Interesting to compare with Pathfinder and MER rovers similar views.

 

Camera system data

Lander Camera

Resolution: 2858x500 px
Bit depth: 6 bits

Filters

IR3, IR2, IR2, Surv
BB1, BB2, BB3, BB4
Red, Green, Blue, Sun