Core Imaging Library (CIL)

The Core Imaging Library (CIL) is a set of modules for each process involved in the data analysis workflow of the CT datasets. Starting from the preparing the dataset for reconstruction which involves applying filters to remove the noise and beam hardening to correct the datasets etc to quantifying the segmented volume. The motivation for creating this library is to provide the CT imaging community with set of tools that is easily accessible and that can be integrated into existing workflows such as SAVU. The algorithms are contributed by the community and the core CCPi staff have reengineered the code to make them run faster, easily accessible and maintainable.

Main repository is at:

Latest release is through conda at:

Software Architecture of CIL

CIL consists of a number of algorithms for pre/post processing, reconstruction, quantification and segmentation. These algorithms are implemented in five modules, namely pre-processing, reconstruction, segmentation, quantification, and volume viewer, as depicted in figure above. It also shows the major input and output data types to the modules. For convenience, CIL also comes with a volume data viewer (optional) to enable ease of exploration volume data. In the first CIL release, our targeted user group is tomography algorithm developers who are interested in exploring and testing CIL algorithms with their existing data. Thus, we make two assumptions about users as follows. They have:

  •  Basic know-how and prior experience of tomography image analysis
  •  Knowledge about how to convert their test data, beyond the one provided in the documentation of CIL, into the numpy array format expected by the CIL modules

Thus, CIL itself doesn’t provide converters to accommodate a wide range of data formats.

In the current release, each module is independent from each other, i.e. there is no dependency between the inputs and outputs of the modules. Developers can use these modules independently to analyse their data.

The documentation for CIL is available at