This document describes the road map for software development related to analysis of imaging data from the instruments at ESS. It is in particular aimed at the needs of the ODIN beamline as this is the main imaging instrument and the first instrument at ESS providing imaging capabilities. The milestones defined in the technical annex are mainly of administrative nature. Therefore, we have defined seven work packages to describe the tasks to solve and the milestones to define when they should be completed. The work packages are ordered according to the time line on the last page.

Lead: A. Kaestner

This work package aims at planning the project and to define the requirements based on the list of features required by the instrument team.

• M1: All documents ready

• T1.1: Write road map
• T1.2: Write quality plan
• T1.3: Write requirement document

• D1.1: Roadmap outlining the work to be done and when it should be completed.
• D1.2: A quality plan describing a quality framework for the development.
• D1.3: A requirement document based on the list provided by the instrument team.

Lead: A. Kaestner

To develop and maintain image processing and analysis that are typically used to work with data from neutron imaging experiments as outlined in the requirements document by the instrument team. This includes image arithmetic, statistics, filter techniques and requirements with respect to CT reconstruction. Basic analysis includes segmentation and classification techniques to define and handle regions of interest. This work will build on the existing code base developed at PSI but will also be supported by other open-source libraries. This is, in particular, relevant for reconstruction algorithms. The development will mainly be based on open-source libraries.

• M2.1: Reconstruction using iterative reconstruction techniques demonstrated.
• M2.2: Image processing demonstrated.

• T2.1: CT reconstruction. [CC]
  To expand and maintain the CT reconstruction pipeline in the framework of MuhRec. This expansion will take two particular directions:
  1. Adding reconstruction algorithms and projection manipulation algorithms (e.g. artefact reduction filters) and (
  2. improving the user-friendliness.
• T2.2: Image processing techniques. [AK]
  Image processing is a wide field and here the focus will be on filter techniques in particular to improve signal to noise ratio and artefact reduction and segmentation techniques and pruning to provide analysis regions for the following analysis in WP3. Image management to crop and rearrange the images into efficient structures for the analysis. The algorithms form the core but may not be easily accessible to the users, therefore image processing shall be implemented in user tools, either interactive or script based.
• T2.3: Handling of regions and binning. [AK]
  Manual and automated segmentation identifies regions that can be used for the analysis. A generalized framework for handling analysis using different regions will be developed. The large data sets over several dimensions (spatial, wavelength, time, etc.) makes it important to provide resampling or binning methods to reduce the data flow to the analysis. This resampling shall be implemented and be able to operate with selected regions of interest.

• D2.1: Iterative reconstruction techniques (e.g. SIRT, DART) are integrated in the reconstruction pipeline.
• D2.2: Image processing tool kit including ROI handling is available for scripting and GUI.

Lead: C. Carminati

This is the central work package and aims at providing tools for the analysis of wavelength resolved image data mainly originating from time of flight experiments. Quantifying information

• M3.1: Different use-cases for Bragg edge analysis have been demonstrated.
• M3.2: Analysis of nGI data using SAS-view has been demonstrated.

• T3.1 Develop and evaluate tools for Bragg edge analysis. [CC]
  Bragg edges can be analyzed in different ways depending on the objectives of the experiment. This ranges from finding the wavelength of a single edge to a full Rietveld refinement to specify the crystallographic parameters from the provided spectrum. This analysis shall be possible per pixel or on regions defined using manual or automated methods.
• T3.2 Visualization of fitted results [CC]
  It shall be possible to visualize parameters from fits in different combinations.
• T3.3 Develop and evaluate tools for beam modulation techniques. [AK]
  Techniques using gratings require a reduction step to provide the information. Modulation sequences using different wavelengths and grating distances provides information that can be analyzed using SANS fitting tools. The existing reduction tool nGITool shall be revised to handle series of data and to provide interoperability with SANS fitting tools like SASView or SASfit.
• T3.4: Develop support for reduction of stroboscopic data. [CC]
  Stroboscopic imaging with ToF requires specific rebinning schemes that use a combination of time stamps, source pulses, and process trigger signals to provide spectra of images for each process position. Support for such rebinning schemes shall be implemented in the image management tool kit.

• D3.1: Bragg edge analysis globally and for single edges.
• D3.2: nGItool with port to SAS-view.
• D3.3: Provide advanced binning techniques for resampling and reshuffling of data from stroboscopic experiments.

Lead: A. Kaestner

Users shall be provided different ways to learn to use the developed analysis tools. The user documentation is central, but also other teaching forms are planned.

• M4.1: A wiki is provided as platform for user documentation.
• M4.2: A concept for publishing tutorials has been developed.
• M4.3: A collection of tutorials are published online.

• T4.1: Provide and maintain online user documentation.
• T4.2: Provide tutorial material.
• T4.3: Organize hands-on trainings on the use of the developed tool suites.

• D4.1: A wiki for the documentation. Updates are made to reflect new features.
• D4.2: Tutorials to explain different features.

Lead: A. Kaestner

This is a technical work package to guarantee that the data can be read, processed, and written within specified time limits. Scripting is an essential ability to allow flexible mass processing and to allow users to add and test new or modified algorithms to the processing workflow.

• M5.1: Accessing experiment data according different binning/resampling schemes.
• M5.2: Central library functionality can be used with python
• M5.3: Analysis can be performed on cluster.

• T5.1: Handling common data file formats. [CC] Data from ToF imaging instruments come in different formats and structures depending on the nature of the experiment. The tool shall be able to open and store image data using the most common formats.
• T5.2: Scripting enabled through python integration. [AK] It is not possible to cover and foresee all variations of analysis in a single application. Python scripting interface allows the user to use existing library functions to refine their analysis and presentation of the results.
• T5.3: Automated mass processing supported by parallel computing. [AK] Imaging instruments and in particular ToF instruments produce massive amounts of data that needs to be analyzed in short time to provide the user with information to make decisions

• D5.1: File formats like NeXus, fits, and TIFF can be read and written by all tools.
• D5.2: Libraries can be accessed by python scripts and python can be used to enhance the functionality of compiled applications.
• D5.3: Analysis tools implemented to run on clusters.

Lead: A. Kaestner

Users should be able to download and run the analysis tools without needing to do extensive manual copying or even compilation. Robustness shall be guaranteed through testing strategies on different levels (from code to application level). All developed code shall be published on a public repository as open source.

• M6.1: Installers are available as nightly builds for all supported OS.
• M6.2: Handover to DMSC.

• T6.1: Provide installers for each supported OS. [CC]
• T6.2: Set up and maintain system for automated building and unit testing. [CC]
  This task is to great extent already implemented but new components will be added to the framework. The unit test coverage of legacy code shall also be improved.
• T6.3: Prepare hand-over to DMSC. [AK]
  At the later stages of the IKC the responsibility of the code shall be transferred to ESS/DMSC.

• D6.1: Deployment routines for supported OS.
• D6.2: Running build system.

Lead: C. Carminati
Objectives: Images provide visual information and a natural first step is

• M7: Software can be purchased.

• T7.1: Scientific visualization software can be purchased.

• D7.1: A report presenting different visualization options, providing a recommendation for purchase.