Wednesday, November 15, 2017

Approaching a Defence in the Lab

The first of our current crop of 5 MScs from 2016 will be defending his thesis in December. This post captures Eric's thoughts as he approaches this milestone and ponders his contributions to planetary science.  While any one thesis is incremental, it is undoubtedly an advance; a stone placed atop what came before that raises the island of science higher.

by Eric Shear

It’s the time of year where I reflect on my research and how far I’ve come. The first draft of my thesis is due at the end of this week, so it’s crunch time for me. This particular project reminds me that science is not all breakthroughs. It’s more usually a series of partial successes and dead ends. It is these roadblocks that help us more, by showing us what doesn’t work. In either case, I must document my research. Perhaps someone else will build on what I’ve learned to build a better spacecraft camera.

Since my last post about using LCDs to increase contrast in spacecraft cameras, I’ve made a great deal of progress. I’ve taken over 90 images of the sun with clouds present in the field of view (but not obscuring the sun). Each image was with at least one LCD, and two-thirds of them were with two LCDs in the optical path. All images were taken with the same exposure time and gain.

The biggest difference I’ve noticed is that the sun is so much more powerful than a mere table lamp, that its rays effortlessly penetrate the darkened patches in the LCDs with little attenuation. Take a look at the trio of photos at the top of this post. At left is the photo taken with one LCD filter, unactivated. At center is the same photo taken with two LCD filters, one activated so the circular block-image is visible. At right is both filters activated with both block-images overlapping to attenuate as much light as possible. There isn’t much of a difference between the centre and right photos.

To drive this point home, I took histograms of each image. Basically, a histogram counts the number of times a value appears in each “bin”. In this case, the bins are the gray values of the pixels, from 0 (pure black) to 255 (pure white). In the first histogram, the largest peak is in the nearly black areas, with a second peak in the dark gray areas. As seen below, the largest shift occurs between the left and centre photos.

To interpret these histograms, it helps to know what each shade of gray represents. The dark gray areas are the sky, the light gray areas are the clouds, and the white areas are the sun itself and its reflections off the clouds. Having a second filter in the optical path, even unactivated, causes the whole image to shift in a lighter direction, with less blacks and more grays. Activating the second filter in the optical path makes very little difference.

Overall, I believe I’ve established that LCD contrast enhancement has very little science return even with a high-contrast LCD. There might be better ways of removing direct sunlight in order to improve contrast and dynamic range, such as directly manipulating the light detector array in the camera. As I’ve mentioned in my last post about this subject, digital camera manufacturers aren’t in the habit of letting their customers do that.

No comments:

Post a Comment