import matplotlib.pyplot as plt

from skimage.measure import shannon_entropy

import numpy

from scipy.stats import kurtosis, skew

from pandas import DataFrame

import numpy as np

from skimage.feature import greycomatrix, greycoprops

import cv2

import sys, os

#Adding a comment, doing some github testing with a new token.

#Testing pycharm push with the new token

#READ HERE: THE pip installs you should run

#Click on terminal in the bottom left.

#pip install matplotlib

#pip install scipy

#pip install skimage

#pip install pandas

#pip install numpy

#pip install opencv-python

sys.stdout = open(os.devnull, 'w') # Suppressing print statements

#Getting the name of the image:

# name='TestImages/testimage_2_ang105width2.tif'

# name ='TestImages/collagen1.tif'

name='TestImages/Picture 7.tif'

# name ='SmallTestImages/C1_0pt4_F_FOV1_REDO_c4_sh.tif'

#Reading the image from a file:

img=cv2.imread(name)

#Converting this image to grayscale, if it isn't already..

img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

# images = imread_collection( 'TestImages/testimage_1_ang90width2.tif', conserve_memory=True)

#Only want to analyze 1 image at a time,

images=[img]

plt.figure(100)

plt.imshow(img,cmap=plt.cm.gray)

plt.show()

def getCorrelation(input_image,angles,dist):

return all_corr

## Setting up an array from 0 to 180

ang_mult= np.arange(180)

# converting this array into radians

angles= [np.pi/180*x for x in ang_mult]

# An empty array which will contain 4 sub-arrays of the correlations at each distance

dist_corr=[]

#TODO: PLAY WITH THESE VALUES: all_dist array contains the distances you would like to analyze

## Code only handles 4 distances right now:

all_dist=[16]

#Running the getCorrelation function at different distance parameters

for i in range(0,len(all_dist)):

correllations= getCorrelation(images[0],angles,all_dist[i])

dist_corr.append(correllations)

#Finding the dominant angle to draw in red on the original image

#Do this by looking at the line corresponding the last (highest) distance .

reported_angle= np.where(dist_corr[len(dist_corr)-1]==np.max(dist_corr[len(dist_corr)-1]))

#If there are multiple angles with the same max correlation, take the median.

if len(reported_angle)>0:

reported_angle=int(np.median(reported_angle))

# The GLCM actually uses angles ina different format, so we convert them..

true_angle=180-reported_angle

print("The true angle is:", true_angle)

#Converting the true angle into radians.

tru_angle_rad=true_angle*np.pi/180

#Plotting the dominant angle line on the image.

norm_vector=[np.sin(tru_angle_rad),np.cos(tru_angle_rad)]

norm_vector=np.array([norm_vector[1],norm_vector[0]*-1])

im=images[0]

h=im.shape[0]

w=im.shape[1]

centre=np.array([int(w/2),int(h/2)])

diag_half=int(w/2)

pt1 = np.subtract(centre, np.multiply(norm_vector, diag_half))

pt2 = np.add(centre, np.multiply(norm_vector, diag_half))

print("Point 1", pt1, "Point 2", pt2)

#Now that we have the two points of the line, we draw it on matplotlib

plt.figure(4)

plt.imshow(im,cmap=plt.cm.gray)

plt.title(f'Line direction, angle={true_angle}')

plt.plot([pt1[0],pt2[0]],[pt1[1],pt2[1]],color='red',label='Calculated line',linewidth=2.0)

plt.legend()

plt.draw()

## Running a plot for the correllations at different distances & angles.

plt.figure(6)

plt.title('Correlation vs angles')

angles_adj_for_plt = [180-y for y in ang_mult]

sys.stdout = sys.__stdout__ # Re-allowing print statements

for dist_index,corr in enumerate(dist_corr):

plt.plot(angles_adj_for_plt,corr, markerfacecolor='black', markersize=3, linewidth=2, label= f"Distance ={all_dist[dist_index]}")

corr_set=len(set(corr))

if len(corr)-corr_set==0:

print("Unique Correlations found at a distance value of",all_dist[dist_index])

plt.ylabel('Correlation')

plt.xlabel('Angle')

plt.legend()

plt.show()