Skip to content

2'. Curating training dataset

Romain F. Laine edited this page Nov 25, 2018 · 1 revision

Performed by: LabelObjects_BATCH.m

This code allows the user to manually label the particles extracted by ExtractObjects_fromImage_BATCH.m into the different classes to create a training dataset.

Input: This requires the output from ExtractObjects_fromImage_BATCH.m.

Paramaters: The border size for the saved single particles images can be set.

Output: The manual classification and the individual particles are saved separately in class folders. If the output from ExtractDescriptors_BATCH.m are present a file can be created with the manual labelling appended.