classification.py

# coding: utf-8

# In[1]:


from sklearn.preprocessing import LabelEncoder
from sklearn.preprocessing import MinMaxScaler
from sklearn.ensemble import RandomForestClassifier
import numpy as np
import mahotas
import cv2
import os
import h5py
import glob
from matplotlib.pyplot import imshow
from matplotlib import pyplot as plt
import io
from PIL import Image
from main import *


# In[2]:


fixed_size = tuple((250, 250))
train_path = "imagedataset/"
num_trees = 100
bins = 8
test_size = 0.10
seed = 9


# In[3]:


def fd_hu_moments(image):
    image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    feature = cv2.HuMoments(cv2.moments(image)).flatten()
    return feature
def fd_haralick(image):
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    haralick = mahotas.features.haralick(gray).mean(axis=0)
    return haralick
def fd_histogram(image, mask=None):
    image = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)
    hist  = cv2.calcHist([image], [0, 1, 2], None, [bins, bins, bins], [0, 256, 0, 256, 0, 256])
    cv2.normalize(hist, hist)
    return hist.flatten()
def Hog_feature(image):
    hog = cv2.HOGDescriptor()
    h = hog.compute(image)
    #print(len(h.flatten()))
    return h.flatten()


# In[4]:


train_labels = os.listdir(train_path)
train_labels.sort()
global_features = []
labels = []


# In[5]:


for training_name in train_labels:
    dir = os.path.join(train_path, training_name)
    current_label = training_name
    for file in os.listdir(dir):
        temp = file.split('.')
        input_file = dir +'/'+file
        #input_file = 'imagedataset/aloo_gobi/70.jpg'
        im = main(0.5 , 8 , 500 , 100 ,input_file)
#         print(input_file)
#         plt.figure()
#         plt.imshow(im)
#         plt.show()        
        
        if(im!=0):
            image = np.array(im)
            fv_hu_moments = fd_hu_moments(image)
            fv_haralick   = fd_haralick(image)
            fv_histogram  = fd_histogram(image)
#             global_feature1 = Hog_feature(image)
            #local_sift = Sift_feature(image)
            global_feature = np.hstack([fv_histogram, fv_haralick, fv_hu_moments])
            labels.append(current_label)
            global_features.append(global_feature)


# In[6]:


targetNames = np.unique(labels)
le = LabelEncoder()
target = le.fit_transform(labels)
scaler = MinMaxScaler(feature_range=(0, 1))
rescaled_features = scaler.fit_transform(global_features)

h5f_data = h5py.File('output/data.h5', 'w')
h5f_data.create_dataset('dataset_1', data=np.array(rescaled_features))

h5f_label = h5py.File('output/labels.h5', 'w')
h5f_label.create_dataset('dataset_1', data=np.array(target))

h5f_data.close()
h5f_label.close()


# In[7]:


from sklearn.model_selection import train_test_split


# In[8]:


(trainDataGlobal, testDataGlobal, trainLabelsGlobal, testLabelsGlobal) = train_test_split(np.array(global_features),np.array(labels),test_size=test_size,random_state=seed)
                                                                                          


# In[9]:


import matplotlib.pyplot as plt
clf  = RandomForestClassifier(n_estimators=206, random_state=16)
count = 0
clf.fit(trainDataGlobal, trainLabelsGlobal)
y_pred = clf.predict(testDataGlobal)
for i in range(len(testLabelsGlobal)):
    if testLabelsGlobal[i] == y_pred[i]:
        count+=1
print(100*count/len(y_pred))


# In[10]:


# test_path = "test"
# for test_name in train_labels:
#     dir = os.path.join(test_path, test_name)
#     for file in os.listdir(dir):
#         temp = file.split('.')
#         address = dir+'/'+file
#         print(address)
#         image = cv2.imread(address)
#         image = cv2.resize(image, fixed_size)
#         fv_hu_moments = fd_hu_moments(image)
#         fv_haralick   = fd_haralick(image)
#         fv_histogram  = fd_histogram(image)
#         local_sift = Sift_feature(image)
#         global_feature = np.hstack([fv_histogram, fv_haralick, fv_hu_moments,local_sift])
#         prediction = clf.predict(global_feature.reshape(1,-1))[0]
#         cv2.putText(image, train_labels[prediction], (20,30), cv2.FONT_HERSHEY_SIMPLEX, 1.0, (0,255,255), 3)
#         plt.imshow(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
#         plt.show()


# In[11]:


from sklearn.linear_model import LogisticRegression
count = 0
logisticRegr = LogisticRegression()
logisticRegr.fit(trainDataGlobal, trainLabelsGlobal)
y_pred = logisticRegr.predict(testDataGlobal)
for i in range(len(testLabelsGlobal)):
    if testLabelsGlobal[i] == y_pred[i]:
        count+=1
print(100*count/len(y_pred))


# In[12]:


from sklearn.svm import SVC  
count = 0
svclassifier = SVC(kernel='linear')  
svclassifier.fit(trainDataGlobal, trainLabelsGlobal)  


# In[13]:


y_pred = svclassifier.predict(testDataGlobal)
for i in range(len(testLabelsGlobal)):
    if testLabelsGlobal[i] == y_pred[i]:
        count+=1
print(100*count/len(y_pred))


# In[14]:


svclassifier = SVC(kernel='poly', degree=8)  
svclassifier.fit(trainDataGlobal, trainLabelsGlobal) 
count = 0
y_pred = svclassifier.predict(testDataGlobal)
for i in range(len(testLabelsGlobal)):
    if testLabelsGlobal[i] == y_pred[i]:
        count+=1
print(100*count/len(y_pred))


# In[15]:


svclassifier = SVC(kernel='rbf')  
svclassifier.fit(trainDataGlobal, trainLabelsGlobal) 
count = 0
y_pred = svclassifier.predict(testDataGlobal)
for i in range(len(testLabelsGlobal)):
    if testLabelsGlobal[i] == y_pred[i]:
        count+=1
print(100*count/len(y_pred))