ComputerVision/ps7.py at main · ICHeavy/ComputerVision · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.patches as patches
from skimage.feature import hog
import os
import cv2
from sklearn.svm import LinearSVC

def getHOGVector(image_name):

  #read the image
  image = plt.imread(image_name)

  #compute the HOG features
  features, hog_image = hog(image,cells_per_block=(2, 2),visualize=True,channel_axis=2)

  #return the features and the corresponding hog image
  return features, hog_image

def getXnY(file):
    xtrain = np.empty(shape=(0, 1188))
    ytrain = np.empty(shape=(0,))
    for image_name in os.listdir('input/p1/'+file+'_imgs'):
        #get the label
        label = int(image_name[0])

        #read the image
        image = plt.imread('input/p1/train_imgs/' + image_name)

        #compute the HOG features
        features, hog_image = hog(image,cells_per_block=(2, 2),visualize=True,channel_axis=2)

        #append
        xtrain = np.append(xtrain, np.array([features]), axis=0)
        ytrain = np.append(ytrain, np.array([label]), axis=0)

    #print the shapes of X_train and y_train
    print(file,' shape:', xtrain.shape)
    print(file, ' shape:', ytrain.shape)
    return xtrain, ytrain

def classifyMe(x, y):
    clf = LinearSVC()
    #train the classifier using the X_train and y_train values
    clf.fit(x,y)
    return clf

def trainModel(clf):
    #loop through the testing images
    k = 1
    for image_name in os.listdir('input/p1/test_imgs'):
        #read the image
        print(image_name, "\n")
        image = plt.imread('input/p1/test_imgs/' + image_name)

        #find the maximum score
        max_score = -1
        max_i = 0
        max_j = 0

        #loop through the image
        for i in range(image.shape[0] - 31):
            for j in range(image.shape[1] - 95):
                #extract the window
                window = image[i:i+32, j:j+96]

                #compute the HOG features
                features, hog_image = hog(window, orientations=9,pixels_per_cell=(8, 8),cells_per_block=(2, 2),visualize=True,
                                        block_norm='L2-Hys', channel_axis=2)

                #get the score
                score = clf.decision_function(np.array([features]))

                #update the maximum score
                if score > max_score:
                    max_score = score
                    max_i = i
                    max_j = j

        #check if the score is greater than 1.2
        if max_score > 1.2:
            #draw a rectangle
            cv2.rectangle(image, (max_j, max_i), (max_j+32, max_i+96), (0, 0, 255), 2)
        fn = 'output/ps7-1-d-'+ str(k) +'.png'
        cv2.imwrite(fn, image)
        k+=1

# 1a
#test this on 'input/p1/car.jpg'
features, hog_img = getHOGVector('input/p1/car.jpg')

#display the hog image
plt.imshow(hog_img)
plt.savefig('ps7-1-a.png')
plt.show()

# 1b
test = "test"
train = "train"
xtrain, ytrain = getXnY(train)

# 1c
classy = classifyMe(xtrain, ytrain)

# 1d
xtest,_ = getXnY(test)
print("almost")
trainModel(classy)