rlemasklib

Library for manipulating masks stored in run-length-encoded format.

This library is an extended version of the pycocotools library’s RLE functions, originally developed by Piotr Dollár and Tsung-Yi Lin for the COCO dataset [1].

There are two ways to use this library:

1. with the RLEMask class, which is an object-oriented way to manipulate RLE masks (recommended)

2. with global functions, which take RLE masks in a dictionary representation, with the keys ‘counts’ and ‘size’

Classes

RLEMask

Run-length encoded mask.

Functions

_set_module_for_docs(module_name, module_globals, ...)

Override __module__ on exported objects so sphinx-codeautolink resolves names.

References

[1]

Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. Microsoft COCO: common objects in context. In European Conference on Computer Vision (ECCV). 2014. URL: https://cocodataset.org, doi:10.1007/978-3-319-10602-1_48.