def descriptors_from_class(dataset, class_img_paths, class_number, option = constants.ORB_FEAT_OPTION):
"""
Gets all the local descriptors for a class. If an image has a side with more than 640 pixels it will be resized
leaving the biggest side at 640 pixels and conserving the aspect ratio for the other side.
Args:
dataset (Dataset object): An object that stores information about the dataset.
class_img_paths (array of strings): The paths for each image in certain class.
class_number (integer): The number of the class.
option (integer): If this is 49 (The key '1') uses ORB features, else use SIFT.
Returns:
numpy float matrix: Each row are the descriptors found in an image of the class
"""
des = None
step = (constants.STEP_PERCENTAGE * len(class_img_paths)) / 100
for i in range(len(class_img_paths)):
img_path = class_img_paths[i]
img = cv2.imread(img_path)
resize_to = 640
h, w, channels = img.shape
if h > resize_to or w > resize_to:
img = utils.resize(img, resize_to, h, w)
if option == constants.ORB_FEAT_OPTION:
des_name = "ORB"
new_des = orb(img)
else:
des_name = "SIFT"
new_des = sift(img)
if new_des is not None:
if des is None:
des = np.array(new_des, dtype=np.float32)
else:
des = np.vstack((des, np.array(new_des)))
# Print a message to show the status of the function
if i % step == 0:
percentage = (100 * i) / len(class_img_paths)
message = "Calculated {0} descriptors for image {1} of {2}({3}%) of class number {4} ...".format(
des_name, i, len(class_img_paths), percentage, class_number
)
print(message)
message = "* Finished getting the descriptors for the class number {0}*".format(class_number)
print(message)
print("Number of descriptors in class: {0}".format(len(des)))
dataset.set_class_count(class_number, len(des))
return des
评论列表
文章目录
