def get_images(url='../static/img', file_extension='.jpg'):
with open(f"{url}/README.txt") as f:
descr = f.read()
filenames = [
f"{url}/{filename}" for filename in sorted(os.listdir(url))
if filename.endswith(file_extension)
]
images = [imread(filename) for filename in filenames]
return Bunch(images=images, filenames=filenames, DESC=descr)
def read_file(filename) -> np.array:
image = np.array(imread(filename))
assert image.shape == (26, 40, 3)
image = image[:, :, 0]
integral_img = integral_image(image)
assert integral_img.shape == (26, 40)
images = [image]
# change images to get more samples
flipped_image = np.fliplr(image)
images.append(integral_image(flipped_image))
flipped_image_tr = np.flipud(image)
images.append(integral_image(flipped_image_tr))
return images
def read_file(fname):
image = imread(fname)
assert image.shape == (26, 40, 3)
return image[:, :, 0]
images_and_predictions = list(zip(digits.images[n_samples // 2:], predictedY))
for index, [image, prediction] in enumerate(images_and_predictions[:5]):
plt.subplot(2, 5, index + 6)
plt.axis('on')
plt.imshow(image, cmap=plt.cm.gray_r, interpolation='nearest')
plt.title('Prediction: %i' % prediction)
print("Original Values: ", digits.target[n_samples // 2:(n_samples // 2) + 5])
plt.show()
#Install Pillow library
#from PIL
#from scipy.misc import imread , imresize, bytescale
from sklearn.externals._pilutil import imresize, imread, bytescale
img = imread("three.jpeg")
img = imresize(img, (8, 8))
classifier = svm.SVC(gamma=0.001)
classifier.fit(imageData[:], digits.target[:])
img = img.astype(digits.images.dtype)
img = bytescale(img, high=16.0, low=0)
print("img.shape : ", img.shape)
print("\n", img)
x_testData = []
for row in img:
for col in row:
x_testData.append(sum(col) / 3.0)
for index, [image, prediction] in enumerate(images_and_predictions[:5]):
plt.subplot(
2, 5, index + 6
) # dividing plotting space into 2 rows and 5 column, and plotting graph in 2nd row
plt.axis('on') # ticks
plt.imshow(image, cmap=plt.cm.gray_r, interpolation='nearest')
plt.title('Prediction : %i' % prediction)
# original target values of images plotted in graph
print("Original Values: ", digits.target[n_samples // 2:(n_samples // 2) + 5])
# plt.show()
# from scipy.misc import imread, imresize, bytescale
# from matplotlib.pyplot import imread
img = imread("FourC.jpeg")
# resizing images to required dimension
img = imresize(img, (8, 8))
classifier = svm.SVC(gamma=0.001)
classifier.fit(imageData[:], digits.target[:])
# making type of images same as earlier(data set of sklearn)
img = img.astype(digits.images.dtype)
img = bytescale(img, high=16.0, low=0)
print("img.shape : ", img.shape)
print("\n", img)
x_testData = []
img_dim = (350, 350)
# show original image
orig_img = cv2.imread("image1.jpg")
cv2.imshow("Original image", orig_img)
# Convert to HSV
hsv_img = cv2.cvtColor(orig_img, cv2.COLOR_BGR2HSV)
cv2.imshow("HSV image", hsv_img)
cv2.waitKey(0)
cv2.destroyAllWindows()
# Gray Scale image
gray_img = cv2.cvtColor(orig_img, cv2.COLOR_RGB2GRAY)
cv2.imwrite("gray.jpg", gray_img)
# Static Threshold images
th_gray = imread("gray.jpg", 0)
ret, th1 = cv2.threshold(th_gray, img_dim[0], img_dim[1], cv2.THRESH_BINARY)
ret, th2 = cv2.threshold(th_gray, img_dim[0], img_dim[1],
cv2.THRESH_BINARY_INV)
ret, th3 = cv2.threshold(th_gray, img_dim[0], img_dim[1], cv2.THRESH_TRUNC)
ret, th4 = cv2.threshold(th_gray, img_dim[0], img_dim[1], cv2.THRESH_TOZERO)
ret, th5 = cv2.threshold(th_gray, img_dim[0], img_dim[1],
cv2.THRESH_TOZERO_INV)
th_names = [
"original image", "Binary", "Binary invert", "Truncate", "ToZero",
"ToZero Invert"
]
img_ths = [th_gray, th1, th2, th3, th4, th5]
labels = [1, 2, 3, 4, 5, 6]
for i in range(6):
plt.pyplot.subplot(2, 3, i + 1),