Cell detection under defocus blur
Type: Bachelor EE/CS/HMI
If you are interested please contact :
As the demand of medical image analysis increases, the aid of deep leaning models has shown its great potential in improving diagnose speed, releasing medical experts from numerous workloads. However, a high-performance medical used model requires not only high accuracy, but also robustness.
For whole-slide cell images, one of the most common corruptions is defocus blur , because the focus of a microscope is inappropriately set, or the positions of cells are non-coplanar. In large-scale cell image analysis, the out-of-focus problem is inevitable and must be dealt with to ensure reliable predictions.
The goal of this project is to detect cells in partial-blurry images (mainly due to defocus blur) and explore the efficiency of defocus blur detection algorithms  in improving robustness towards such corruption.