Metadata-Version: 2.1
Name: sputummrcnn
Version: 0.9
Summary: Custom Mask RCNN
Home-page: UNKNOWN
License: UNKNOWN
Platform: UNKNOWN
Requires-Python: >=3.6
License-File: LICENSE.txt
Requires-Dist: numpy
Requires-Dist: scipy
Requires-Dist: Pillow
Requires-Dist: cython
Requires-Dist: matplotlib
Requires-Dist: pandas
Requires-Dist: scikit-image
Requires-Dist: tensorflow (==1.15.0)
Requires-Dist: keras (==2.2.4)
Requires-Dist: opencv-python
Requires-Dist: h5py (==2.10.0)
Requires-Dist: imgaug
Requires-Dist: theano
Requires-Dist: IPython[all]

Sputum MaskRCNN
================

A Custom Python Package for Implementing MaskRCNN. This is a private package intended 
for implementing Mask RCNN Algorithm on Sputum Project, its ML Model generation 
and related resources. 

Installing
----------
.. note::
	It is advised to install in a virtual enviroment

run : pip install sputummrcnn

Features
--------
* Mask RCNN Implementation on Sputum Files.
	Given the set of images and the annotation files the package trains the models 
	as per user requirements. The Images can then be predicted or cropped using 
	the other features in the package.

* Feature Extraction.
	Based on the Cropped Images, certain features are extracted by the package,
	and then saved as a parquet file which can be used as training dataset for
	the Machine Learning model.

* Machine Learning Model & Analysis
	After Feature files are generated, based on Random Forest or K Nearest Neighbour
	Algorithm a Machine Learning Model is trained based on user preference. It's
	accuracy and confusion is also shown so as to select the best one.

Credits
-------

https://github.com/matterport/Mask_RCNN

