InsightSoftwareConsortium / InsightSoftwareConsortium/ITK

Modify the `Thresholding` module's tests parameters to provide meaningul output images

Open
#1,737 2 comments 0 reactions 0 assignees View on GitHub
area:Filtering status:Use_Milestone_Backlog type:Testing
Dominant language
C++
Stars
1.7k
Forks
748
Avg merge
1d 1h
Merged PRs (30d)
64

Description

### Description

Modify the tests in the `Thresholding` module so that they have appropriate parameters so as to provide a meaningful thresholded output images: e.g. `itkOtsuThresholdImageFilterTestShort` generates a thresholded output which is not really meaningful/useful (no anatomical or meaningful feature is thresholded).

### Steps to Reproduce

1. Configure and build ITK, including the `ITKData` folder.
2. **Identify** the tests in the `Thresholding` module that do not provide meaningful images, i.e. tests whose images in the `Baseline` folder are not anatomically meaningful (e.g. they contain a single value, no anatomical or meaningul structure from the input image can be distinguished, etc.).
3. **Modify the test parameters** so that meaningful output data is generated..

### Expected behavior

Regression tests should have appropriate parameters to yield meaningful outputs for a proper check of the behavior of the filter.

### Actual behavior

Some tests do not have appropriate parameters that allow to check if a meaningful/appropriate/expected thresholding is taking place.

### Reproducibility

100%.

### Versions

`master`

### Environment

All environments.

### Additional Information

See [this comment](https://github.com/InsightSoftwareConsortium/ITK/pull/511#issuecomment-605536928) and the context in PR #511.

@richardbeare your insight is highly appreciated.

Contributor guide

Open the contributing guide

Research direction

Start by building ITK with the ITKData folder and inspect the Thresholding module tests, including itkOtsuThresholdImageFilterTestShort, alongside their images in the Baseline folder. Identify outputs that are single-valued or lack meaningful anatomical structure, then adjust the test parameters and verify that the resulting regression images provide useful thresholding checks.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
computer-vision, testing
Issue type
Refactor
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.