def main(args):
imgs_list, imgs_dir = get_image_list(args.image_path)
if args.trimap_path is None:
trimaps_list = None
else:
trimaps_list, _ = get_image_list(args.trimap_path)
predictor = Predictor(args)
predictor.run(imgs=imgs_list, trimaps=trimaps_list, imgs_dir=imgs_dir)
def recognize(self,
img,
horizontal_list=None,
free_list=None,
reformat=True,
imgH=32):
if reformat:
img, img_cv_grey = reformat_input(img_cv_grey)
if (horizontal_list == None) and (free_list == None):
b, y_max, x_max = img.shape
ratio = x_max / y_max
max_width = int(imgH * ratio)
crop_img = cv2.resize(img, (max_width, imgH),
interpolation=Image.ANTIALIAS)
image_list = [([[0, 0], [x_max, 0], [x_max, y_max],
[0, y_max]], crop_img)]
else:
image_list, max_width = get_image_list(horizontal_list,
free_list,
img,
model_height=imgH)
result = get_text(self.recognizer, image_list)
return result
def predict(args, predictor):
input_names = predictor.get_input_names()
input_tensor = predictor.get_input_handle(input_names[0])
output_names = predictor.get_output_names()
output_tensor = predictor.get_output_handle(output_names[0])
test_num = 500
test_time = 0.0
if not args.enable_benchmark:
# for PaddleHubServing
if args.hubserving:
img_list = [args.image_file]
# for predict only
else:
img_list = get_image_list(args.image_file)
for idx, img_name in enumerate(img_list):
if not args.hubserving:
img = cv2.imread(img_name)[:, :, ::-1]
assert img is not None, "Error in loading image: {}".format(
img_name)
else:
img = img_name
inputs = utils.preprocess(img, args)
inputs = np.expand_dims(inputs, axis=0).repeat(args.batch_size,
axis=0).copy()
input_tensor.copy_from_cpu(inputs)
predictor.run()
output = output_tensor.copy_to_cpu()
classes, scores = utils.postprocess(output, args)
if args.hubserving:
return classes, scores
print("Current image file: {}".format(img_name))
print("\ttop-1 class: {0}".format(classes[0]))
print("\ttop-1 score: {0}".format(scores[0]))
else:
for i in range(0, test_num + 10):
inputs = np.random.rand(args.batch_size, 3, 224,
224).astype(np.float32)
start_time = time.time()
input_tensor.copy_from_cpu(inputs)
predictor.run()
output = output_tensor.copy_to_cpu()
output = output.flatten()
if i >= 10:
test_time += time.time() - start_time
time.sleep(0.01) # sleep for T4 GPU
fp_message = "FP16" if args.use_fp16 else "FP32"
trt_msg = "using tensorrt" if args.use_tensorrt else "not using tensorrt"
print("{0}\t{1}\t{2}\tbatch size: {3}\ttime(ms): {4}".format(
args.model, trt_msg, fp_message, args.batch_size,
1000 * test_time / test_num))
def main(args):
env_info = get_sys_env()
place = 'gpu' if env_info['Paddle compiled with cuda'] and env_info[
'GPUs used'] else 'cpu'
paddle.set_device(place)
if not args.cfg:
raise RuntimeError('No configuration file specified.')
cfg = Config(args.cfg)
val_dataset = cfg.val_dataset
if val_dataset is None:
raise RuntimeError(
'The verification dataset is not specified in the configuration file.'
)
elif len(val_dataset) == 0:
raise ValueError(
'The length of val_dataset is 0. Please check if your dataset is valid'
)
msg = '\n---------------Config Information---------------\n'
msg += str(cfg)
msg += '------------------------------------------------'
logger.info(msg)
model = cfg.model
transforms = val_dataset.transforms
image_list, image_dir = get_image_list(args.image_path)
if args.trimap_path is None:
trimap_list = None
else:
trimap_list, _ = get_image_list(args.trimap_path)
logger.info('Number of predict images = {}'.format(len(image_list)))
predict(model,
model_path=args.model_path,
transforms=transforms,
image_list=image_list,
image_dir=image_dir,
trimap_list=trimap_list,
save_dir=args.save_dir)
def recognize(self, img_cv_grey, horizontal_list=None, free_list=None,\
decoder = 'greedy', beamWidth= 5, batch_size = 1,\
workers = 0, allowlist = None, blocklist = None, detail = 1,\
paragraph = False,\
contrast_ths = 0.1,adjust_contrast = 0.5, filter_ths = 0.003,\
reformat=True):
if reformat:
img, img_cv_grey = reformat_input(img_cv_grey)
if (horizontal_list == None) and (free_list == None):
y_max, x_max = img_cv_grey.shape
ratio = x_max / y_max
max_width = int(imgH * ratio)
crop_img = cv2.resize(img_cv_grey, (max_width, imgH),
interpolation=Image.ANTIALIAS)
image_list = [([[0, 0], [x_max, 0], [x_max, y_max],
[0, y_max]], crop_img)]
else:
image_list, max_width = get_image_list(horizontal_list,
free_list,
img_cv_grey,
model_height=imgH)
if allowlist:
ignore_char = ''.join(set(self.character) - set(allowlist))
elif blocklist:
ignore_char = ''.join(set(blocklist))
else:
ignore_char = ''.join(set(self.character) - set(self.lang_char))
if self.model_lang in [
'chinese_tra', 'chinese_sim', 'japanese', 'korean'
]:
decoder = 'greedy'
result = get_text(self.character, imgH, int(max_width), self.recognizer, self.converter, image_list,\
ignore_char, decoder, beamWidth, batch_size, contrast_ths, adjust_contrast, filter_ths,\
workers, self.device)
if self.model_lang == 'arabic':
direction_mode = 'rtl'
result = [list(item) for item in result]
for item in result:
item[1] = get_display(item[1])
else:
direction_mode = 'ltr'
if paragraph:
result = get_paragraph(result, mode=direction_mode)
if detail == 0:
return [item[1] for item in result]
else:
return result
def main(args):
imgs_list, imgs_dir = get_image_list(args.image_path)
if args.trimap_path is None:
trimaps_list = None
else:
trimaps_list, _ = get_image_list(args.trimap_path)
if use_auto_tune(args):
tune_img_nums = 10
auto_tune(args, imgs_list, tune_img_nums)
predictor = Predictor(args)
predictor.run(imgs=imgs_list, trimaps=trimaps_list, imgs_dir=imgs_dir)
if use_auto_tune(args) and \
os.path.exists(args.auto_tuned_shape_file):
os.remove(args.auto_tuned_shape_file)
if args.benchmark:
predictor.autolog.report()
def main():
args = utils.parse_args()
# assign the place
place = paddle.set_device('gpu' if args.use_gpu else 'cpu')
net = architectures.__dict__[args.model](class_dim=args.class_num)
load_dygraph_pretrain(net, args.pretrained_model, args.load_static_weights)
image_list = get_image_list(args.image_file)
for idx, filename in enumerate(image_list):
img = cv2.imread(filename)[:, :, ::-1]
data = utils.preprocess(img, args)
data = np.expand_dims(data, axis=0)
data = paddle.to_tensor(data)
net.eval()
outputs = net(data)
if args.model == "GoogLeNet":
outputs = outputs[0]
outputs = F.softmax(outputs)
outputs = outputs.numpy()
probs = postprocess(outputs)
top1_class_id = 0
rank = 1
print("Current image file: {}".format(filename))
for idx, prob in probs:
print("\ttop{:d}, class id: {:d}, probability: {:.4f}".format(
rank, idx, prob))
if rank == 1:
top1_class_id = idx
rank += 1
if args.pre_label_image:
save_prelabel_results(top1_class_id, filename,
args.pre_label_out_idr)
return
basedir = args.base_dir
input_dir = os.path.join(basedir, args.input_dir)
disp_dir = os.path.join(basedir, args.disp_dir)
if not os.path.exists(disp_dir):
os.mkdir(disp_dir)
if args.save_undistort:
left_dir = os.path.join(basedir, 'left')
right_dir = os.path.join(basedir, 'right')
if not os.path.exists(left_dir):
os.mkdir(left_dir)
if not os.path.exists(right_dir):
os.mkdir(right_dir)
img_list = utils.get_image_list(input_dir)
img_size = cv2.imread(os.path.join(input_dir, img_list[0])).shape[:2][::-1]
img_size = (img_size[0] // 2, img_size[1])
roi = (0, 0, img_size[0], img_size[1])
min_disp = args.min_disp
if args.num_disp is None:
num_disp = (img_size[0] // 8 + 15) & -16
else:
num_disp = args.num_disp
matcher = StereoMatcher(args.calib, img_size)
matcher.create_matcher(min_disp,
num_disp,
args.blocksize,
args.mode,
left_dir = os.path.join(basedir, args.left_dir)
right_dir = os.path.join(basedir, args.right_dir)
debug_dir = os.path.join(basedir, args.debug_dir)
if args.debug and not os.path.exists(debug_dir):
os.mkdir(debug_dir)
pattern_points = np.zeros((np.prod(PATTERN_SIZE), 3), np.float32)
pattern_points[:, :2] = np.indices(PATTERN_SIZE).T.reshape(-1, 2)
pattern_points *= args.square_size
obj_points = []
img_points_l = []
img_points_r = []
image_list = get_image_list(left_dir)
pre_img_l, pre_img_r = None, None
for i, fname in enumerate(image_list):
print('[%d/%d] Processing % s ...' % (i + 1, len(image_list), fname))
img_l = cv2.imread(os.path.join(left_dir, fname))
img_r = cv2.imread(os.path.join(right_dir, fname))
if np.array_equal(img_l, pre_img_l) and np.array_equal(
img_r, pre_img_r):
continue
pre_img_l, pre_img_r = img_l.copy(), img_r.copy()
corners_l = find_chess_board_corners(img_l, PATTERN_SIZE)
corners_r = find_chess_board_corners(img_r, PATTERN_SIZE)
image_size = img_l.shape[:2][::-1]
import os
import sys
import cv2
from utils import get_image_list, split_stereo_image
INPUT_DIR_NAME = 'rgb'
LEFT_DIR_NAME = 'left'
RIGHT_DIR_NAME = 'right'
if __name__ == '__main__':
basedir = sys.argv[1]
input_dir = os.path.join(basedir, INPUT_DIR_NAME)
left_dir = os.path.join(basedir, LEFT_DIR_NAME)
right_dir = os.path.join(basedir, RIGHT_DIR_NAME)
if not os.path.exists(left_dir):
os.mkdir(left_dir)
if not os.path.exists(right_dir):
os.mkdir(right_dir)
image_list = get_image_list(input_dir)
for i, fname in enumerate(image_list):
print("[%d/%d] Spliting % s ..." % (i + 1, len(image_list), fname))
im = cv2.imread(os.path.join(input_dir, fname))
im_l, im_r = split_stereo_image(im)
cv2.imwrite(os.path.join(left_dir, fname), im_l)
cv2.imwrite(os.path.join(right_dir, fname), im_r)
def recognize(self, img_cv_grey, horizontal_list=None, free_list=None,\
decoder = 'greedy', beamWidth= 5, batch_size = 1,\
workers = 0, allowlist = None, blocklist = None, detail = 1,\
rotation_info = None,\
paragraph = False,\
contrast_ths = 0.1,adjust_contrast = 0.5, filter_ths = 0.003,\
reformat=True):
if reformat:
img, img_cv_grey = reformat_input(img_cv_grey)
if allowlist:
ignore_char = ''.join(set(self.character) - set(allowlist))
elif blocklist:
ignore_char = ''.join(set(blocklist))
else:
ignore_char = ''.join(set(self.character) - set(self.lang_char))
if self.model_lang in ['chinese_tra', 'chinese_sim']:
decoder = 'greedy'
if (horizontal_list == None) and (free_list == None):
y_max, x_max = img_cv_grey.shape
horizontal_list = [[0, x_max, 0, y_max]]
free_list = []
# without gpu/parallelization, it is faster to process image one by one
if ((batch_size == 1) or (self.device == 'cpu')) and not rotation_info:
result = []
for bbox in horizontal_list:
h_list = [bbox]
f_list = []
image_list, max_width = get_image_list(h_list,
f_list,
img_cv_grey,
model_height=imgH)
result0 = get_text(self.character, imgH, int(max_width), self.recognizer, self.converter, image_list,\
ignore_char, decoder, beamWidth, batch_size, contrast_ths, adjust_contrast, filter_ths,\
workers, self.device)
result += result0
for bbox in free_list:
h_list = []
f_list = [bbox]
image_list, max_width = get_image_list(h_list,
f_list,
img_cv_grey,
model_height=imgH)
result0 = get_text(self.character, imgH, int(max_width), self.recognizer, self.converter, image_list,\
ignore_char, decoder, beamWidth, batch_size, contrast_ths, adjust_contrast, filter_ths,\
workers, self.device)
result += result0
# default mode will try to process multiple boxes at the same time
else:
image_list, max_width = get_image_list(horizontal_list,
free_list,
img_cv_grey,
model_height=imgH)
image_len = len(image_list)
if rotation_info and image_list:
image_list = make_rotated_img_list(rotation_info, image_list)
max_width = max(max_width, imgH)
result = get_text(self.character, imgH, int(max_width), self.recognizer, self.converter, image_list,\
ignore_char, decoder, beamWidth, batch_size, contrast_ths, adjust_contrast, filter_ths,\
workers, self.device)
if rotation_info and (horizontal_list + free_list):
result = set_result_with_confidence(result, image_len)
if self.model_lang == 'arabic':
direction_mode = 'rtl'
result = [list(item) for item in result]
for item in result:
item[1] = get_display(item[1])
else:
direction_mode = 'ltr'
if paragraph:
result = get_paragraph(result, mode=direction_mode)
if detail == 0:
return [item[1] for item in result]
else:
return result
import os
import argparse
import cv2
import utils
parser = argparse.ArgumentParser()
parser.add_argument('image_dir',
metavar='IMAGE_DIR',
help='the directory of images')
parser.add_argument('--fps',
'-f',
type=int,
default=20,
help='the video playback speed')
args = parser.parse_args()
if __name__ == '__main__':
img_list = utils.get_image_list(args.image_dir)
for (i, fname) in enumerate(img_list):
print('Frame [%d/%d] % s ...' % (i + 1, len(img_list), fname))
img = cv2.imread(os.path.join(args.image_dir, fname))
cv2.imshow('image', img)
cv2.waitKey(int(1000 / args.fps))
cv2.waitKey(0)