def plot_histogram(image):
hist, bins = np.histogram(image.flatten(), 256, [0, 256])
cdf = hist.cumsum()
cdf_normalized = cdf * hist.max() / cdf.max()
plt.plot(cdf_normalized, color='b')
plt.hist(image.flatten(), 256, [0, 256], color='r')
plt.xlim([0, 256])
plt.legend(('cdf', 'histogram'), loc='upper left')
plt.show()
return
def load_data(path):
"""
TODO: Add documentation
:param path:
:return:
"""
labels = []
X = []
y = []
products = os.listdir(path)
for i, product in enumerate(products):
print("Product {} [{}/{}] loaded...".format(product, i + 1,
len(products)))
labels.append(product)
images = os.listdir(os.path.join(path, product))
for filename in images:
img = Image.open(os.path.join(path, product, filename))
img.thumbnail((50, 50), Image.ANTIALIAS)
image = np.array(img)
image = image.flatten()
X.append(image)
y.append(i)
X = np.array(X)
y = np.array(y)
return X, y, labels
def prepare_image(img):
from keras.preprocessing.image import img_to_array
image = img_to_array(img)
# Scale the image pixels by 255 (or use a scaler from sklearn here)
image = 255 - (image)
# Flatten into a 1x28*28 array
image = image.flatten().reshape(-1, 28 * 28)
return (image)
def get_next_batch(batch_size=64):
batch_x = np.zeros([batch_size, IMAGE_HEIGHT * IMAGE_WIDTH])
batch_y = np.zeros([batch_size, MAX_CAPTCHA * CHAR_SET_LEN])
for i in range(batch_size):
name, image = get_name_and_image()
batch_x[i, :] = 1 * (image.flatten())
batch_y[i, :] = name2vec(name)
return batch_x, batch_y
def prepare_image(image):
if image.mode != "RGB":
image = image.convert("RGB")
image_size = (28, 28)
image = image.resize(image_size)
# image.save("fromform.png")
image = img_to_array(image)[:, :, 0]
image /= 255
image = 1 - image
return image.flatten().reshape(-1, 28 * 28)
def FrameClustering(FramePath):
train_data = []
train_labels_one_hot =[]
trainfileList =[]
for root, dirs, files in os.walk(FramePath):
for file in files:
if file.endswith(".jpg"):
fullName = os.path.join(root, file)
trainfileList.append(fullName)
for imagePath in trainfileList:
image = cv2.imread(imagePath)
image = cv2.resize(image, (28, 28))
image = img_to_array(image)
image.ndim
image = image.flatten()
image /= 255
train_data.append(image)
#train_data /= 255
print(image.shape)
print(image.dtype)
dfFrame = pd.DataFrame.from_records(train_data)
dfFrame.shape
#Kmeans to label poses
kmeans = KMeans(n_clusters=12)
kmeansoutput = kmeans.fit(dfFrame)
clusLabel = kmeans.predict(dfFrame)
cluCenter = kmeans.cluster_centers_
Lables = pd.DataFrame(clusLabel)
dfFrame['Cluster'] = clusLabel
#PCA for visulaisation
pca = PCA(n_components=2)
principalComponents = pca.fit_transform(dfFrame)
principalDf = pd.DataFrame(data = principalComponents, columns = ['PCA1', 'PCA2'])
pl.scatter(principalDf['PCA1'],principalDf['PCA2'], c=kmeansoutput.labels_,s=100)
return dfFrame