Sparse imaging and reconstruction in a pixel array

Several applications like Cryo-electron microscopy (Cryo-EM) involve reconstructing sparse binary images. The conventional imaging system for reading out images from a pixel array uses a brute-force pixel-by-pixel readout. This approach is slow and inefficient, especially for a high-resolution imager, motivating the need for new and faster readout systems.

This project aims at realizing faster readout architectures and novel reconstruction algorithms by exploiting the underlying sparse structure of the binary image. The main technological idea behind the project is that the frame rate for sparse imaging can be increased significantly if the pixels are read out in groups and decoded together. Furthermore, in cryo-EM-related applications, it is known that the electron detection probability is high in some regions of the pixel array compared to the other areas (the probability typically vanishes at the edges of the pixel array). The knowledge of the prior statistics can further reduce the required number of readouts, achieving higher frame rates. The project involves the new probabilistic group testing model that enables the actual sparse imaging.

Requirements

For this project, we are looking for a master's student in either electrical engineering or any related study. Furthermore, we are looking for a student who has a background in signal processing, sensors, basic statistical techniques, data analysis, and programming skills in Matlab, Python, and/or C/C++. Strong communication (written and verbal) skills in English are mandatory.

Contact

Last modified: 2024-03-04