Skip to content

Improve threshold selection method to be more automated #8

@tobyallwood

Description

@tobyallwood

Currently thresholds are chosen by looping through possible thresholds and choosing the one producing the most grains. This causes a number of possible issues:

  • The lower and upper bounds of thresholds to search through are entered manually in the config. If the ideal threshold for an image is outside the chosen bounds it will not be selected
  • The step is also manually chosen in the config (step being the amount of change between each checked threshold) so if a small step is chosen the final threshold chosen will be more accurate but it can take a very long time to find. This is also partly due to grain finding being processed for every threshold tested.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions