The Science and Engineering of Digital Photography

(3 Credit Hours) · Undergraduate

Images have become so common in our lives that we often take for granted the immense amount of science and engineering that is required every time we take a photo. From collecting light, to converting photons to electrons, to processing the raw data from a sensor, capturing an image for framing or posting to Instagram is a triumph of decades of discovery and engineering. The goal of this course is to introduce students to basic concepts and applications of optics, optoelectronics, and image processing through digital photography. Students will gain skills in data analysis, critical thinking, and experiment design. This course will help students make connections between the classroom and real-world experiences and observations by challenging them to think deeply about the technology they carry in their hand. In my course, students will be introduced to fundamental concepts of light, optics, semiconductor-based image sensors, image quality, and image processing—all at an introductory level. Student learning will be aided by lectures, in-class problem sets, take-home problem sets, and exploratory hand-on experiences using digital single lens reflex (dSLR) cameras. These hands-on experiences are an integral part of this course and will be designed to have students look at images differently—from the perspective of an engineer. For example, when capturing a picture of the Milky Way above the Sligo coastline in northwest Ireland, how can we quantify the noise on the image sensor and how does it affect image quality? How can this noise be reduced by the user and what are the limits of this approach? The students will use the dSLRs to capture images, analyze them in Matlab or Python, and report their findings. We will post many of these images online (Twitter, online gallery) to share with family and friends back at home. When appropriate, I will supplement these experiences using data from my nanophotonics laboratory. For example, how does the sensitivity and noise of a CMOS sensor in a dSLR compare to a single photon detector or a cryogenically-cooled CMOS sensor? Why might one need better detectors in their respective disciplines?