Add files via upload

Slow_Students è il main (completo) che usa le funzioni Inizialization_images, minimum_analysis e frequency_analysis

Image_Analysis e Image_Inizialization insieme sono equivalenti a Slow_Students ma permettono di fare separatamente la applicazione della maschera scrivendo il risultato su file (30 min).

PROBLEMI:
1- background, dove sbaglio?
2- la riga 61 del file minimum_analysis mi dà errore (credo sia dovuto al fatto che j è definito come una lista, non ho capito bene cosa stia facendo...) Marco puoi aiutarmi?

Author: araihC

Slow_Students-master/Inizialization_images.py

@@ -0,0 +1,123 @@
+
+#-------------------------------FUNCTION DEFINITION-----------------------------
+def Make_a_rectangoular_crop(img_collection, topx, topy, bottomx, bottomy):
+
+    #import matplotlib.pyplot as plt
+    from PIL import Image
+
+    img_collection_cropped = []
+    #img = ((100,100))
+
+    for i in range(len(img_collection)):
+        img = img_collection[i]
+
+        cropped  = img[topx:bottomx, topy:bottomy]
+        img_collection_cropped.append(cropped)
+        #plt.matshow(cropped)
+        #plt.show()
+
+    return(img_collection_cropped)
+
+
+#--------------------------Contours plotting------------------------------------
+
+
+def drawShape(img, coordinates, color):
+        #This function drows all contours founded on the mask
+        # In order to draw our line in red
+    img = skimage.color.gray2rgb(img)
+
+        # Make sure the coordinates are expressed as integers
+    coordinates = coordinates.astype(int)
+
+    img[coordinates[:, 0], coordinates[:, 1]] = color
+
+    return img
+
+
+def Contours_printing(img_float, contours):
+        # Display the image and plot all contours found
+    fig, ax = plt.subplots()
+    ax.imshow(img_float, interpolation='nearest', cmap=plt.cm.gray)
+
+    for n, contour in enumerate(contours):
+        ax.plot(contour[:, 1], contour[:, 0], linewidth=2)
+
+    ax.axis('image')
+    ax.set_xticks([])
+    ax.set_yticks([])
+    plt.show()
+
+        #Create a Black mask of the same dimentions of the image
+    mask = np.zeros_like(img_float) # Create mask where white is what we want, black otherwise
+
+    #Drow contours on 'mask'
+    for contour in contours:
+        mask = drawShape(mask, contour, [255, 0, 0])
+
+#------------------------Find best contour--------------------------------------
+
+def Find_Contours(img):
+
+    import numpy as np
+    from PIL import Image
+    import skimage
+    from skimage import data, io, filters, measure, img_as_float, img_as_uint
+    from matplotlib import pyplot as plt
+
+    Contour_Limit = 0.05
+
+    contour = []
+
+    while (len(contour) != 1):
+        contour = measure.find_contours(img, Contour_Limit)
+        Contour_Limit += 0.001
+        #print Contour_Limit
+
+    return contour
+
+def Find_a_mask(img_collection):
+    #This function finds and drows contures of an image given.
+    #It returns the cropped image as a numpy array of float
+
+        # Find contours the best Contour_Limit for the first image of the serie
+    contours = Find_Contours(img_collection[0])
+
+    return(contours)
+
+#-------------------------------------------------------------------------------
+
+
+def Inside_outside_check(point, contours):
+    #Mi raccomando Contours lo si deve passare come Contours[0]
+
+    from shapely.geometry import Point
+    from shapely.geometry.polygon import Polygon
+
+    point = Point(point)
+    polygon = Polygon(contours)
+    return(polygon.contains(point))
+
+#-------------------------------------------------------------------------------
+
+def Apply_a_mask(img_collection, Contours, DIM_X, DIM_Y):
+
+    import numpy as np
+    #import matplotlib.pyplot as plt
+
+    img_collection=np.asarray(img_collection)
+
+    #print len(img_collection)
+
+    for t in range(len(img_collection)):
+
+        for x in range(DIM_X):
+            for y in range(DIM_Y):
+                point = (x,y)
+                if not(Inside_outside_check(point, Contours[0])):
+                    #se il punto e' nel contorno
+                    img_collection[t,x,y] = 'nan'
+        #plt.matshow(img_collection[t])
+        #plt.show()
+
+    return img_collection

Slow_Students-master/Slow_Waves.py

@@ -0,0 +1,132 @@
+#===============================================================================
+#                                       MAIN
+#===============================================================================
+
+import os
+import numpy as np
+from pylab import *
+from scipy.signal import butter, lfilter, filtfilt, freqz
+from scipy.optimize import curve_fit
+import matplotlib.pyplot as plt
+
+import skimage
+from skimage import data, io, filters, measure
+from skimage import img_as_float, img_as_uint
+
+from Inizialization_images import drawShape, Make_a_rectangoular_crop, Find_a_mask, Apply_a_mask, Inside_outside_check
+from frequency_analysis import bandpass_filter, multiplot
+from minimum_analysis import find_minimum, min_interpol, parab_fit, min_analysis
+
+#------------------------------PARAMETER DEFINITION-----------------------------
+
+# Filter requirements.
+order = 6
+fs = 25.0           # sample rate, Hz
+lowcut = 0.3        # desired cutoff frequency of the filter, Hz
+highcut = 2.5
+
+SAMPLING_TIME = 40. # ms (unit)
+DIM_X = 100         #image dimension
+DIM_Y = 100
+
+MACRO_PIXEL_DIM = 2 #pixel dimention of our new 'macro-pixel'
+
+ZOOM = 4            #minimum fit parameter
+t_min = 0           #time range for minimum research
+t_max = 2
+
+#-----------------------------IMAGE LOADING-------------------------------------
+
+path = "/Users/Mac/Desktop/Uni/LabII/data/t1/provevideo"
+
+
+img_path_list = []
+for i in range(1, 4):
+    img_path = '1_' + str(i) + '.tif'
+    filename = os.path.join(skimage.data_dir, path + img_path)
+
+    img_path_list.append(filename)
+
+# Load all the collection of the images
+img_collection_float = img_as_float(io.imread_collection(img_path_list))
+
+'''
+DA RICONTROLLARE IL BACKGROUND
+img_background = np.zeros((100, 100), np.float64)
+
+# Evaluate the background of the images as the mean over the whole set
+for i in img_collection_float:
+    # Convert all images to float, range -1 to 1, to avoid dangerous overflows
+    img_background += i
+
+img_background /= len(img_collection_float)
+
+# Substract from each image the background
+for i in img_collection_float:
+    i -= img_background
+'''
+
+img_collection_UP = []
+img_collection_DOWN = []
+
+img_collection_UP = Make_a_rectangoular_crop(img_collection_float, 0, 0, 19, 100)
+img_collection_DOWN = Make_a_rectangoular_crop(img_collection_float, 20, 0, 100, 100)
+
+
+#-----------------------Analysis of Down images
+DIM_X=80
+DIM_Y=100
+
+Contours_DOWN = Find_a_mask(img_collection_DOWN)
+
+'''
+#DA VELOCIZZARE QUESTO PUNTO
+'''
+img_collection_DOWN = Apply_a_mask(img_collection_DOWN, Contours_DOWN, DIM_X, DIM_Y)
+
+
+# Now we need to reduce the noise from the images by performing a spatial smoothing
+img_collection_reduced = []
+img_collection_reduced = measure.block_reduce(img_collection_DOWN, (1, MACRO_PIXEL_DIM, MACRO_PIXEL_DIM), np.mean)
+print img_collection_reduced.shape
+
+
+
+#--------------------------------Spectrum analysis------------------------------
+
+# Next we take the fourier transformation of the input along the temporal axis
+img_spectrum = np.fft.rfftn(img_collection_reduced, axes = [0])
+img_spectrum_freq = np.fft.rfftfreq(img_collection_reduced.shape[0], d = 1. / SAMPLING_TIME)
+
+# We take the mean of the spectrum
+spectrum_mean = measure.block_reduce(img_spectrum, (1, np.size(img_collection_reduced,1), np.size(img_collection_reduced,2)), np.mean)
+
+xmax = np.amax(img_collection_reduced[:, 5, 7])
+xmin = np.amin(img_collection_reduced[:, 5, 7])
+
+plt.plot(img_collection_reduced[300:800, 5, 7])
+#plt.show()
+
+#Clean the signal through a bandpass-pass filter
+
+clean_signals = bandpass_filter(img_collection_reduced)
+
+#To have an idea we choose randomly the pixel 5,7
+plt.plot(clean_signals[20, 20,:])
+#plt.show()
+
+#plt.hist(clean_signals[5, 7, :], bins='auto', range = (xmin , xmax))
+#plt.yscale('log')
+#plt.show()
+
+
+#Multiplot of different pixels
+#plt.plot = multiplot(10, 40, 10, 10, 40, 10, 650, 750)
+#plt.show()
+
+
+#----------------------------------MINIMUM ANALYSIS-------------------------------
+
+#Identify min and max values
+minimum_signal = []
+minimum_signal = min_analysis(clean_signals, ZOOM, t_min, t_max)

Slow_Students-master/analisi_separata/Image_analysis.py

@@ -0,0 +1,86 @@
+#===============================================================================
+#                                       MAIN
+#===============================================================================
+
+import os
+import numpy as np
+from pylab import *
+from scipy.signal import butter, lfilter, filtfilt, freqz
+from scipy.optimize import curve_fit
+import matplotlib.pyplot as plt
+
+import skimage
+from skimage import data, io, filters, measure
+from skimage import img_as_float, img_as_uint
+
+from Inizialization_images import drawShape, Make_a_rectangoular_crop, Find_a_mask, Apply_a_mask, Inside_outside_check
+from frequency_analysis import bandpass_filter, multiplot
+from minimum_analysis import find_minimum, min_interpol, parab_fit, min_analysis
+
+#------------------------------PARAMETER DEFINITION-----------------------------
+
+# Filter requirements.
+order = 6
+fs = 25.0           # sample rate, Hz
+lowcut = 0.3        # desired cutoff frequency of the filter, Hz
+highcut = 2.5
+
+SAMPLING_TIME = 40. # ms (unit)
+DIM_X_REDUCED = 40         #image dimension
+DIM_Y_REDUCED = 50
+
+MACRO_PIXEL_DIM = 2 #pixel dimention of our new 'macro-pixel'
+
+ZOOM = 4            #minimum fit parameter
+t_min = 0           #time range for minimum research
+t_max = 2
+
+#-----------------------------IMAGE LOADING-------------------------------------
+
+Images2D = np.loadtxt('inizialized_images.txt')
+print Images2D.shape
+print Images2D
+
+img_collection_reduced = np.reshape(Images2D, ((1000, DIM_X_REDUCED, DIM_Y_REDUCED)),order='C')
+print img_collection_reduced.shape
+#plt.matshow(img_collection_reduced[0])
+#plt.show()
+
+#--------------------------------Spectrum analysis------------------------------
+
+# Next we take the fourier transformation of the input along the temporal axis
+img_spectrum = np.fft.rfftn(img_collection_reduced, axes = [0])
+img_spectrum_freq = np.fft.rfftfreq(img_collection_reduced.shape[0], d = 1. / SAMPLING_TIME)
+
+# We take the mean of the spectrum
+spectrum_mean = measure.block_reduce(img_spectrum, (1, np.size(img_collection_reduced,1), np.size(img_collection_reduced,2)), np.mean)
+
+xmax = np.amax(img_collection_reduced[:, 5, 7])
+xmin = np.amin(img_collection_reduced[:, 5, 7])
+
+plt.plot(img_collection_reduced[300:800, 5, 7])
+#plt.show()
+
+#Clean the signal through a bandpass-pass filter
+
+clean_signals = bandpass_filter(img_collection_reduced)
+
+#To have an idea we choose randomly the pixel 5,7
+plt.plot(clean_signals[20, 20,:])
+#plt.show()
+
+#plt.hist(clean_signals[5, 7, :], bins='auto', range = (xmin , xmax))
+#plt.yscale('log')
+#plt.show()
+
+
+#Multiplot of different pixels
+#plt.plot = multiplot(10, 40, 10, 10, 40, 10, 650, 750)
+#plt.show()
+
+
+#----------------------------------MINIMUM ANALYSIS-------------------------------
+
+#Identify min and max values
+minimum_signal = []
+minimum_signal = min_analysis(clean_signals, ZOOM, t_min, t_max)