Statismo is an open source framework for statistical shape modeling. It supports all shape modeling tasks, from model building to shape analysis, and makes it easy to exchange models with your peers.
Statismo features a unique approach to shape model building using kernel methods. This makes it possible to build expressive shape models, even if you have only a few example shapes available.
Shape modeling is tough. Decades of research have led to sophisticated but complex algorithms that can make shape modeling unwildy. Statismo reduces the complexity by focusing on a single but very powerful concept which unifies non-rigid registration of shapes, images and shape model fitting.
Statistical models can be studied from many different point of views. Statismo treats statistical models as probability distributions over the modeled objects. The statismo API is formulated in terms of operations on probability distributions and hides technical details of the actual implementation from the user.
Ideally, shape models are build once and used in many differnet applications. Statismo makes it easy to exchange models. By adding information about the data and parameters that have been used during model building, statismo makes sure that it is possible to know how the model has been built.
Each application of shape models has specific requirements and may use a specific toolkit to represent shapes or images. Statismo provides an abstraction which makes it possible to use statismo together with any image or shape representation provided by your favorite tookit.
We plan to release a first version as open source shortly.