PDE Software for Digital Image Management

摘要

In Phase I research, we investigated and developed image segmentation and registration algorithms using curvature-based flow-driven nonlinear partial differential equation imaging models and algorithms, including total variation, level-set, and active contour techniques. We implemented the developed algorithms in C and S+, an open and extensible software environment, to allow us to do rapid prototyping and evaluation of the methods. We applied combinations of the methods developed to two applications in medical imaging, prostate segmentation and brain segmentation and registration.