Department of Computer Science
Diversity Programs in Engineering
CURIE Academy 2015
Design Project: The Science and the Magic of Photos
Prof. Noah Snavely
Upson Hall • July 12–18, 2015
The CURIE Academy is a one-week summer residential program for high school girls who excel in math and science, enjoy solving problems, and want to learn more about careers in engineering. CURIE scholars spend their mornings learning about the various fields within engineering, and spend their afternoons working on a design project. This year's design project experience is titled "The Science and the Magic of Photos" and it will be led by Prof. Noah Snavely from the Department of Computer Science.
Photos are Serious Business. Every day, billions of photos are captured and shared by people across the globe. These photos help us capture important memories, and communicate key moments to our friends and family. But photos are also very important for science and engineering, whether it is for understanding the health of a forest from a satellite image, building a 3D model of a city from aerial imagery, measuring the size of a glacier, or diagnosing health problems from X-rays or other types of images.
However, photos are limited in a number of ways: they are difficult to capture in dark places, they can be blurry or out of focus, and they only capture a small view of the world. In the last decade, as photos have migrated from prints in shoeboxes and physical photo albums, to digital files on our computers and mobile devices, we have started seen computation play a key role in trying to make photos better: to turn normal photos into photos with “superhuman” vision, with less blur, a larger field of view, or even the ability to change the objects that are in focus after the photo is captured. New types of sensors have also given us new types of photos, such as images that also record the 3D shape of the scene. These advances are made possible by new algorithms that can analyze and improve photos even as they are captured. This area is called computational photography, and it is impacting the photos we capture every day, as well as enabling new ways of seeing the world.
The 2015 Curie Scholars will explore computational photography by creating algorithms for computational photography that can solve a number of problems with conventional photography. The week-long project will begin with three laboratory sessions that will use the science toolbox MATLAB to explore photography. In the first lab, scholar will learn the basics of representing and transforming images as sets of pixels, and regenerate the first color photographs ever taken by the Russian photographer Prokudin-Gorskii in the early 1900s. In the second lab, scholars will learn how to filter images to count the number of cells in a microscope image or turn photos into cartoon-style works of art. In the third lab, scholars will take photos from space and shrink them into much smaller size with the same quality, and enhance badly garbled satellite images. After these laboratory sessions, scholars will work in small groups on a photography project of their choosing. Example projects include monitoring heart rate from webcam video, creating “hybrid” images that show different objects depending on how close the viewer is, and creating “tilt-shift” miniature images.