if __name__ == '__main__':
symbols = []
#creating desc for each file with label
for (dirpath,dirnames,filenames) in os.walk(path):
for dirname in sorted(dirnames):
print(dirname)
symbols.append(dirname)
for(direcpath,direcnames,files) in os.walk(path+"/"+dirname):
file_list = list(files)
#random.shuffle(file_list)
for file in file_list[:BATCH_SIZE]:
actual_path=path+"/"+dirname+"/"+file
frame = cv2.imread(actual_path)
des = func(frame)
if des is None: # Eliminates empty items.
continue
img_descs.append(des)
y.append(label)
label=label+1
#finding indexes of test train and validate
y=np.array(y)
training_idxs, test_idxs, val_idxs = train_test_val_split_idxs(len(img_descs), 0.4, 0.0)
#creating histogram using kmeans minibatch cluster model
X, cluster_model = cluster_features(img_descs, training_idxs, MiniBatchKMeans(n_clusters=150))
#splitting data into test, train, validate using the indexes
X_train, X_test, X_val, y_train, y_test, y_val = perform_data_split(X, y, training_idxs, test_idxs, val_idxs)
print("mlp started")
clf.fit(X_train, y_train)
y_pred = clf.predict(X_test)
calc_accuracy("MLP classifier", y_test, y_pred)
#creating desc for each file with label
for (dirpath, dirnames, filenames) in os.walk(path):
for dirname in dirnames:
print(dirname)
print(path + "/" + dirname)
for (direcpath, direcnames, files) in os.walk(path + "/" + dirname):
for file in files:
actual_path = path + "/" + dirname + "/" + file
print(actual_path)
des = func(actual_path)
if des != None:
img_descs.append(des)
y.append(label)
label = label + 1
#finding indexes of test train and validate
y = np.array(y)
training_idxs, test_idxs, val_idxs = train_test_val_split_idxs(
len(img_descs), 0.4, 0.0)
#creating histogram using kmeans minibatch cluster model
X, cluster_model = cluster_features(img_descs, training_idxs,
MiniBatchKMeans(n_clusters=150))
#splitting data into test, train, validate using the indexes
img = cv2.flip(img, 1)
cv2.imshow("original", img)
axis_length = min(img.shape[0], img.shape[1])
diff = abs(img.shape[1] - img.shape[0])
if img.shape[0] < img.shape[1]:
new_img = img[:, diff // 2:img.shape[1] - diff // 2]
else:
new_img = img[diff // 2:img.shape[0] - diff // 2, :]
resized_img = cv2.resize(new_img, (200, 200))
image_data = cv2.imencode('.jpg', resized_img)[1].tostring()
with open('input.jpg', 'wb') as image_file:
image_file.write(image_data)
img_des = func(resized_img)
try:
X, cluster_model = cluster_features([img_des], range(1),
MiniBatchKMeans(n_clusters=150))
#print(cluster_model)
y_pred = clf.predict(X)
print("\n\nPredicted symbol:")
print(symbols[int(y_pred)])
except ValueError:
print("less features")
if cv2.waitKey(2500) == ord('q'):
break
# Following line should... <-- This should work fine now