Merge from upstream (#28)

* Update particle_analysis.rst

* Add 3D plotting of properties

* Add tests for 3D plotting

* Update test_plotting.py

* Update test_plotting.py

* Update ptcl_class.py

* Update 3D plot label setting

* Fix plotting bug

* Update ptcl_class.py

* Update test_plotting.py

* Add 3D plotting to docs

* Update 3D plotting

* Add additional 3d plotting functionality

* Bugfix for area unit calculation

* added Niblack and Sauvola segmentation

Added two additional local segmentation methods: Niblack and Sauvola.

Also converted local threshold method to odd integers only, as specified by skimage.filters.threshold_local

* Bugfix for Invert check in SegUI

Fixed invert toggle not applying to SegUI image after 'Update'.

Fixed typo in ParticleAnalysis
A typo in docs
Changes to segmentation.rst to add Niblack and Sauvola.

changed parameter files back.

* Revert "Bugfix for Invert check in SegUI"

This reverts commit 3e65293.

* no altered parameters files

commit back to 3e65293 without broken parameters files

* Assitional clustering tests

added new clustering test capable of sets of parameters for each thresholding method

* changed test-clustering

* fixed test particle clustering

* more Parameter testing

test particle clustering extended to all thresholding methods apart from sauvola.

* Local Filter kernel Fixes

disables option to change local filter kernel until a local segmentation method is selected.

For Niblack, Sauvola and Local thresholding methods local kernel filter is locked to odd numbers

updated parameters file to account for new local filter kernel values

* test for scikit image update

* remove parameters in regionprops

* Update test_parameters.dat

* updated test_parameters

* updated watershed function used

* Updated watershed in SegUI

* Update test_params.dat again

* corrected clean_borders

corrected clean borders to take place after watershed separation to exclude less complete particles

* improved compatibility for scikits

reverted scikit-learn earliest to 0.21 and scikit-image to 0.17.1

* updated test parameters

updated test parameters in test_parameters.dat.

updated the default parameters as these were overriding the user-created parameters if they were set at zero

* Update setup.py

* adds new clustering methods

added new clustering methods for individual hyperspy objects and series of hyperspy objects

* Update ParticleAnalysis.py

* Update test_parameters.dat

* Update test_particle_clustering.py

* more filter parameters

* Update ParticleAnalysis.py

* New SegUI tab

added new SegUI tab to develop improved labelling methods

* added Radion buttons for different tools

* added line tool

* Update SegUI.py

* fixed tool alignment

* Fixed Tool selection

* added polygon tool

* Updated coordinates of SegUI to use QPoints

* Tidied Tools and line drawings

* Fixed flood Fill

flood fill no longer darkens touside of image and correctly fills enclosed shapes

* arr flipped to ARGB

* updated colour selection of flood fill

* moved flood fill to separate function

* added test for learn clustering

* added documentation and test fix

* updated tests

* New SegUI tab

added new SegUI tab to develop improved labelling methods

* added Radion buttons for different tools

* added line tool

* Update SegUI.py

* fixed tool alignment

* Fixed Tool selection

* added polygon tool

* Updated coordinates of SegUI to use QPoints

* Tidied Tools and line drawings

* Fixed flood Fill

flood fill no longer darkens touside of image and correctly fills enclosed shapes

* arr flipped to ARGB

* updated colour selection of flood fill

* moved flood fill to separate function

* fixed color flood filling and added color tool buttons

* flood fill colour is pen dependent

* fixed flooding

* added clear button

* updated saving labels

* added trainable segmentation

* updated SegUI layout

* corrected trainables

* updated classifier

* tidied SegUI

* commented ClusterTrained

* Working trainable segmentation

* Update ParticleAnalysis.py

* Update labels to v0.5.0

* added difference of gaussians

* added membrane projection kernel

* added documentation

* added training to SegUI

* fixed labels displaying on canvas

* Update SegUI.py

* fixed clustering methods

* Adds docs and moves functions

Moved all functions related to clustering and classifying to a new file 'segimgs.py'

* rewrites clustering algorithms

* moves segimgs.py to correct folder

* improves docs & fixes bugs

* adds partial test and bugfixes

* adds test for trainSeg

* Update test_particle_clustering.py

* removes saving mask to png

* fixes tests

* updates docs

updates docs with example of classifying multiple images

* Update segmentation.rst

* Update test_particle_clustering.py

Co-authored-by: CameronGBell <cameronbell2236@gmail.com>
Co-authored-by: CameronGBell <70440659+CameronGBell@users.noreply.github.com>

Author: TomSlater

ParticleSpy/Data/trainingmask.png

@@ -8,13 +8,15 @@
 from ParticleSpy.segptcls import process
 import numpy as np
 from ParticleSpy.ptcl_class import Particle, Particle_list
+from ParticleSpy.custom_kernels import membrane_projection
 from skimage import filters, morphology
 from skimage.measure import label, regionprops, perimeter
+from sklearn import preprocessing
+from sklearn.cluster import DBSCAN, KMeans
 import ParticleSpy.find_zoneaxis as zone
 import warnings
 import h5py
 import inspect
-import matplotlib.pyplot as plt
 import pandas as pd
 import trackpy
 import os
@@ -35,8 +37,8 @@ def ParticleAnalysis(acquisition,parameters,particles=None,mask=np.zeros((1))):
         List of already analysed particles that the output can be appended
         to.
     mask: Numpy array
-        Numpy array of same 2D size as acquisition that contains a mask of presegmented
-        particles.
+        Numpy array of same 2D size as acquisition that contains a mask of 
+        presegmented particles.
         
     Returns
     -------
@@ -170,7 +172,7 @@ def ParticleAnalysisSeries(image_series,parameters,particles=None):
     
     Parameters
     ----------
-    image_series: Hyperpsy signal object or list of hyperspy signal objects.
+    image_series: Hyperspy signal object or list of hyperspy signal objects.
         Hyperpsy signal object containing nanoparticle images or a list of signal
          objects that contains a time series.
     parameters: Dictionary of parameters
@@ -231,7 +233,7 @@ def timeseriesanalysis(particles,max_dist=1,memory=3,properties=['area']):
 
     t = trackpy.link(df,max_dist,memory=memory)
     return(t)
-    
+
 def store_image(particle,image,params):
     ii = np.where(particle.mask)