transforms

Transform Catalog

The below table has all of the transformations currently available in the Data Transform widget. Each transform should have a test.cpp file (Column 2). The test.cpp file should include tests to make sure it can be populated by a JSON template (e.g. it can be run by loading data with a JSON file). Some transforms will “just work” because they are parameterless, but transforms with parameters will need to have them also described with a JSON link and put in the transform parameter factory. Documentation for users should be provided to describe the algorithm and its uses. Finally, a benchmark.cpp file should be created and placed in the same directory.

Transform Name	Has .test.cpp?	JSON-based Test?	Parameters in ParameterFactory?	User Docs Exist?	Has benchmark.cpp?	Benchmark Results
Calculate Mask Centroid	Yes	Yes	Yes	Yes
Convert Mask to Line	Yes	Yes	Yes	No	No
Skeletonize Mask	Yes	Yes	Yes	Yes	No
Remove Small Connected Components	Yes	Yes	Yes	Yes	No
Apply Median Filter	Yes	Yes	Yes	Yes	No
Calculate Area	Yes	Yes	Yes	Yes	No
Fill Mask Holes	Yes	Yes	Yes	Yes	No
Calculate Mask Principal Axis	Yes	Yes	Yes	Yes	No
Whisker Tracing	No	N/A	No	No	No
Clip Line by Reference Line	Yes	Yes	Yes	Yes	No
Line Alignment to Bright Features	Yes	Yes	Yes	No	Yes	10 images in 2ms
Calculate Line to Point Distance	Yes	Yes	Yes	Yes	No
Resample Line	Yes	Yes	Yes	Yes	No
Extract Line Subsegment	Yes	Yes	Yes	Yes	No
Calculate Line Angle	Yes	Yes	Yes	Yes	No
Calculate Line Curvature	Yes	Yes	Yes	Yes	No
Extract Point from Line	Yes	Yes	Yes	Yes	No
Filter	Yes	No	No	No	No
Hilbert Phase	Yes	Yes	Yes	Yes	No
Threshold Event Detection	Yes	Yes	Yes	Yes	No
Threshold Interval Detection	Yes	Yes	Yes	Yes	No
Scale and Normalize	Yes	Yes	Yes	No	No
Group Intervals	Yes	Yes	Yes	Yes	No