Image processing trick: Removing interline transfer smear from Curiosity photos
Over the weekend, Curiosity took a set of 75 photos with her arm-mounted MAHLI camera, enough to produce a spectacular new self-portrait of the rover sitting at the Kimberley, prepared to drill at Windjana. I absolutely love this new self-portrait; it contains everything you'd want. There's the rover, Mount Sharp, beautiful rover tracks, beautiful outcrop, a mound in the middle distance, and the walls of Gale crater quite clear in the background. They even shot it with the head in two different positions, so you can choose to assemble it either with the head facing you or gazing down at the outcrop. I am pretty sure some version of it will end up on the cover of my book.
Curiosity sol 613 MAHLI self-portrait: component images
Curiosity's arm-mounted MAHLI camera took 75 individual photos in order to cover the entire rover (with mast head in two different positions) as well as the east side of the Kimberley field site and Mount Sharp in the background.
I'm not going to post a fully-assembled version here yet, because it's still a work in progress. I want to give the amateur -- and professional! -- community more time to produce what I think will be the definitive version. In the meantime, here is a partial version that I made for my desktop:
NASA / JPL / MSSS / Emily Lakdawalla
Curiosity sol 613 MAHLI self-portrait: desktop background
Curiosity captured this view of itself at the Kimberley on sol 613 (April 27, 2014), as it prepared to drill at Windjana.
On the interline transfer CCDs used in many Point Grey cameras, each pixel consists of a light sensitive area, called the photosensor, and a light-shielded area, called the vertical transfer register. The vertical transfer register is used to shift the charge out of the photosensor and off of the sensor. Although vertical transfer cells are approximately 100,000 times less sensitive than the light sensitive region, photons may still tunnel into this area, especially when exposed to very bright light. This tunneling, or leaking, is the source of smear.
During image readout, image pixels are shifted vertically downard, row by row, through the vertical transfer cells to the readout register. If there is any leakage of charge into an area of the vertical transfer register, it will be picked up and shifted downward, so that every pixel above and below the area will contain this extra charge. The result is a relatively bright vertical stripe across the entire image.
I was having a hard time putting an explanation in words, so I just decided to record it. It worked even though I had to keep telling my computer to stop worrying about not having enough resources. Enjoy!
An explainer on using Photoshop to remove interline transfer smear from MAHLI or Mastcam image frames.
Note: one step you may not get in the tutorial is how I link the adjustment layer just to the layer I'm subtracting from the original. After you add an adjustment layer, Control-click on the line between the adjustment layer and the layer immediately below it, and that will make the adjustment layer apply only to the layer immediately below it rather than to the entire image, as would be the default.