Are you interested in digital photography or imaging sensors? We are exploring ways of improving the digital imaging sensors used in cameras. Our group is focused on improving digital camera performance and lifetime. Unfortunately, digital imagers like any other microelectronic devices, develop defects over time. Unlike other microchips, most in-field defects in digital imagers begin appearing soon after fabrication, are permanent, and their number increases continuously over the lifetime of the sensor. These faulty pixels degrade the quality of the image generated by the sensor. Although the impact of defects can be overcome by factory recalibration, this is often expensive and even infeasible for imagers used in many applications (eg those in remote locations). We have been investigating imager in-field defect development for several years now and have identified the characteristics and rate of faulty pixel development. 

We are working on 3 areas: the identification of defects as cameras age, creating new ways to recover the missing pixel information, and the testing of new sensor designs. As sizes of imaging sensors become larger both in pixel count and area, the possibility of pixel defects increases during manufacturing, and over the lifetime of the sensor. People do not want to throw away expensive cameras just because they have dead pixels in it, but find such dead spots annoying in pictures. We are exploring ways of correcting this using both software which knows the defect’s characteristics and potentially fault tolerant pixels. Students will help us test, analyze and simulate the response of this design.  Previous students have also been part of published conference papers on these results and part of it has resulted in a patent application. 

Links to more information

Want to learn more about defects in cameras – go to the the Image Sensor overview paper for a more complete introduction.:

Depending on the student’s background this project would range from:
(1) Experimental testing of digital cameras to identify and evaluate defects.  This can include hardware development for the testing the cameras in the lab, and software development to run the tests (controlling the cameras).  You will gain experience in fully understanding the operation of digital cameras (how the sensor transforms the raw image into a final picture). 
(2) Developing software programs to analyze the image data to locate the defects, and extract their parameters. This will involve expanding our software which detects digital defects both from laboratory tests and from regular digital images. Statistical analysis of the data will allow us to identify the causal source of the defects, relates defects growth to camera parameters such as pixel size, and suggest ways to improve imaging sensor lifetimes. Hence you gain an idea of how experimental design and statistical data analysis is applied for real research.
(3) Help develop algorithms and software for recovering the true image hidden by the defects.  You will be using experimental methods to compare the original true image, captured by a sensor without damage, to what an algorithm projects as the corrected image.  This is applying programming skills (Matlab or C) to real world image applications.
(4) Potentially working to develop programs that will use digital images posted on the internet to detect the presence of defects in those cameras.  We want to gather information from images on the web that help us identify defects in those images.  Potentially creating a citizen science web site so the public can be part of this.

Previous summer students have also been part of published conference papers on these results (including one that is part of a patent application),

Skills Needed:
Student should be in third year or above.  Some combination of the following skills are needed, but not all are required .  The skill set will determine the type of project. 
(1) A background in digital photography and a general liking of experimental work, but. you need not be a serious amateur photographer. This work involves testing cameras ranging from high end DSLRs to cell phones. 
(2) Good computer skills for PC based systems: spreadsheets & Matlab and/or C programming.
(3) Taken an introductory Optics courses from physics.
(4) Experience with adobe photoshop is useful though not required.
(5) Electronics or Physics background.
(6) Taken a statistics course .