Segmentation is a fundamental aspect of image processing with many important applications, such as contouring in medical imaging. The variational approach has been widely used to partition 2D and 3D data in numerous practical settings, but effectively incorporating specific knowledge of the target object(s) continues to be a significant challenge. This minisymposium will address recent advances in variational methods for segmentation, considering a wide range of techniques. The work is primarily motivated by associated applications; a number of which which will be discussed in detail.

Interactive Variational Segmentation in Medical Imaging

Jack Spencer (University of Liverpool)

Joint CT Reconstruction and Segmentation with Discriminative Dictionary Learning

Yiqiu Dong (Technical University of Denmark)

Segmentation-Based Blind Image Restoration of Space-Variant Defocus Blur

Leah Bar (Tel-Aviv University and maxq.ai )

Minimal Paths and Geodesic Metrics for Image Segmentation and Tubular Structure Extraction

Da Chen (University Paris Dauphine, PSL Research University)

Organizers:

Jack Spencer (University of Liverpool)

Keywords:

computed tomography, computer vision, image reconstruction, image segmentation, machine learning