def verify_annotation(img_path, ann_path, labels):
img = cv2.imread(img_path)
boxes = [] # xywh
with open(ann_path, 'r') as f:
for line in f:
line = line[:-1].split()
line[0] = int(line[0])
line[1:] = [float(i) for i in line[1:]]
boxes.append(line)
boxes = np.asarray(boxes)
boxes[:, 1:] = xywh2xyxy(boxes[:, 1:]) # xyxy
if boxes[0, 1] <= 1:
boxes[:, 1::2] *= img.shape[1]
boxes[:, 2::2] *= img.shape[0]
colors = [[random.randint(0, 255) for _ in range(3)] for _ in range(len(labels))]
for box in boxes:
label = '%s' % (labels[int(box[0])])
plot_one_box(box[1:], img, label=label, color=colors[int(box[0])])
return img
def __init__(self, path, img_size=416, batch_size=16, augment=False, hyp=None, rect=False, image_weights=False,
cache_labels=False, cache_images=False, single_cls=False):
path = str(Path(path)) # os-agnostic
assert os.path.isfile(path), 'File not found %s. See %s' % (path, help_url)
with open(path, 'r') as f:
self.img_files = [x.replace('/', os.sep) for x in f.read().splitlines() # os-agnostic
if os.path.splitext(x)[-1].lower() in img_formats]
n = len(self.img_files)
assert n > 0, 'No images found in %s. See %s' % (path, help_url)
bi = np.floor(np.arange(n) / batch_size).astype(np.int) # batch index
nb = bi[-1] + 1 # number of batches
self.n = n
self.batch = bi # batch index of image
self.img_size = img_size
self.augment = augment
self.hyp = hyp
self.image_weights = image_weights
self.rect = False if image_weights else rect
# Define labels
self.label_files = [x.replace('images', 'labels').replace(os.path.splitext(x)[-1], '.txt')
for x in self.img_files]
# Rectangular Training https://github.com/ultralytics/yolov3/issues/232
if self.rect:
# Read image shapes (wh)
sp = path.replace('.txt', '.shapes') # shapefile path
try:
with open(sp, 'r') as f: # read existing shapefile
s = [x.split() for x in f.read().splitlines()]
assert len(s) == n, 'Shapefile out of sync'
except:
s = [exif_size(Image.open(f)) for f in tqdm(self.img_files, desc='Reading image shapes')]
np.savetxt(sp, s, fmt='%g') # overwrites existing (if any)
# Sort by aspect ratio
s = np.array(s, dtype=np.float64)
ar = s[:, 1] / s[:, 0] # aspect ratio
i = ar.argsort()
self.img_files = [self.img_files[i] for i in i]
self.label_files = [self.label_files[i] for i in i]
self.shapes = s[i] # wh
ar = ar[i]
# Set training image shapes
shapes = [[1, 1]] * nb
for i in range(nb):
ari = ar[bi == i]
mini, maxi = ari.min(), ari.max()
if maxi < 1:
shapes[i] = [maxi, 1]
elif mini > 1:
shapes[i] = [1, 1 / mini]
self.batch_shapes = np.ceil(np.array(shapes) * img_size / 32.).astype(np.int) * 32
# Preload labels (required for weighted CE training)
self.imgs = [None] * n
self.labels = [None] * n
if cache_labels or image_weights: # cache labels for faster training
self.labels = [np.zeros((0, 5))] * n
extract_bounding_boxes = False
create_datasubset = False
pbar = tqdm(self.label_files, desc='Caching labels')
nm, nf, ne, ns, nd = 0, 0, 0, 0, 0 # number missing, found, empty, datasubset, duplicate
for i, file in enumerate(pbar):
try:
with open(file, 'r') as f:
l = np.array([x.split() for x in f.read().splitlines()], dtype=np.float32)
except:
nm += 1 # print('missing labels for image %s' % self.img_files[i]) # file missing
continue
if l.shape[0]:
assert l.shape[1] == 5, '> 5 label columns: %s' % file
assert (l >= 0).all(), 'negative labels: %s' % file
assert (l[:, 1:] <= 1).all(), 'non-normalized or out of bounds coordinate labels: %s' % file
if np.unique(l, axis=0).shape[0] < l.shape[0]: # duplicate rows
nd += 1 # print('WARNING: duplicate rows in %s' % self.label_files[i]) # duplicate rows
if single_cls:
l[:, 0] = 0 # force dataset into single-class mode
self.labels[i] = l
nf += 1 # file found
# Create subdataset (a smaller dataset)
if create_datasubset and ns < 1E4:
if ns == 0:
create_folder(path='./datasubset')
os.makedirs('./datasubset/images')
exclude_classes = 43
if exclude_classes not in l[:, 0]:
ns += 1
# shutil.copy(src=self.img_files[i], dst='./datasubset/images/') # copy image
with open('./datasubset/images.txt', 'a') as f:
f.write(self.img_files[i] + '\n')
# Extract object detection boxes for a second stage classifier
if extract_bounding_boxes:
p = Path(self.img_files[i])
img = cv2.imread(str(p))
h, w = img.shape[:2]
for j, x in enumerate(l):
f = '%s%sclassifier%s%g_%g_%s' % (p.parent.parent, os.sep, os.sep, x[0], j, p.name)
if not os.path.exists(Path(f).parent):
os.makedirs(Path(f).parent) # make new output folder
b = x[1:] * [w, h, w, h] # box
b[2:] = b[2:].max() # rectangle to square
b[2:] = b[2:] * 1.3 + 30 # pad
b = xywh2xyxy(b.reshape(-1, 4)).ravel().astype(np.int)
b[[0, 2]] = np.clip(b[[0, 2]], 0, w) # clip boxes outside of image
b[[1, 3]] = np.clip(b[[1, 3]], 0, h)
assert cv2.imwrite(f, img[b[1]:b[3], b[0]:b[2]]), 'Failure extracting classifier boxes'
else:
ne += 1 # print('empty labels for image %s' % self.img_files[i]) # file empty
# os.system("rm '%s' '%s'" % (self.img_files[i], self.label_files[i])) # remove
pbar.desc = 'Caching labels (%g found, %g missing, %g empty, %g duplicate, for %g images)' % (
nf, nm, ne, nd, n)
assert nf > 0, 'No labels found. See %s' % help_url
# Cache images into memory for faster training (WARNING: large datasets may exceed system RAM)
if cache_images: # if training
gb = 0 # Gigabytes of cached images
pbar = tqdm(range(len(self.img_files)), desc='Caching images')
self.img_hw0, self.img_hw = [None] * n, [None] * n
for i in pbar: # max 10k images
self.imgs[i], self.img_hw0[i], self.img_hw[i] = load_image(self, i) # img, hw_original, hw_resized
gb += self.imgs[i].nbytes
pbar.desc = 'Caching images (%.1fGB)' % (gb / 1E9)
# Detect corrupted images https://medium.com/joelthchao/programmatically-detect-corrupted-image-8c1b2006c3d3
detect_corrupted_images = False
if detect_corrupted_images:
from skimage import io # conda install -c conda-forge scikit-image
for file in tqdm(self.img_files, desc='Detecting corrupted images'):
try:
_ = io.imread(file)
except:
print('Corrupted image detected: %s' % file)
img_paths = img_paths[restart_index:]
ann_paths = ann_paths[restart_index:]
# Labels
with open(os.path.join(DS_PATH, 'diff.names'), 'w') as f:
for n in NEW_CLS:
f.write('%s\n' % n)
for ip, ap in zip(img_paths, ann_paths):
# Read original annotation
img = cv2.imread(ip)
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
boxes = get_ann_boxes(ap)
np_boxes = np.asarray(boxes)
np_boxes[:, 1:] = xywh2xyxy(np_boxes[:, 1:])
np_boxes[:, 1:] = np.clip(np_boxes[:, 1:], 0, 1) # IMPORTANT
np_boxes[:, 1:] = scale_xyxy(np_boxes[:, 1:], img.shape[:2])
irt_np_boxes = np.insert(np_boxes, 1, 1, axis=1)
# Info output
print(('%20s: %s' * 2) % ('Image', os.path.basename(ap)[:-4] + '.jpg',
'Hands', np_boxes.shape[0]))
# Differentiation
lr_boxes = diff_hands(img,
irt_np_boxes,
extend_scale=0,
color_thres=10,
bin_thres=5,
max_region=False,
def __init__(self, path, img_size=416, batch_size=16, augment=False, rect=True, image_weights=False):
with open(path, 'r') as f:
img_files = f.read().splitlines()
self.img_files = list(filter(lambda x: len(x) > 0, img_files))
n = len(self.img_files)
bi = np.floor(np.arange(n) / batch_size).astype(np.int) # batch index
nb = bi[-1] + 1 # number of batches
assert n > 0, 'No images found in %s' % path
self.n = n
self.batch = bi # batch index of image
self.img_size = img_size
self.augment = augment
self.image_weights = image_weights
self.rect = False if image_weights else rect
# Define labels
self.label_files = [x.replace('images', 'labels') for x in self.img_files]
for f in img_formats:
self.label_files = [x.replace(f, '.txt') for x in self.label_files]
# Rectangular Training https://github.com/ultralytics/yolov3/issues/232
if self.rect:
from PIL import Image
# Read image shapes
sp = 'data' + os.sep + path.replace('.txt', '.shapes').split(os.sep)[-1] # shapefile path
if os.path.exists(sp): # read existing shapefile
with open(sp, 'r') as f:
s = np.array([x.split() for x in f.read().splitlines()], dtype=np.float32)
assert len(s) == n, 'Shapefile out of sync, please delete %s and rerun' % sp
else: # no shapefile, so read shape using PIL and write shapefile for next time (faster)
s = np.array([Image.open(f).size for f in tqdm(self.img_files, desc='Reading image shapes')])
np.savetxt(sp, s, fmt='%g')
# Sort by aspect ratio
ar = s[:, 1] / s[:, 0] # aspect ratio
i = ar.argsort()
self.img_files = [self.img_files[i] for i in i]
self.label_files = [self.label_files[i] for i in i]
ar = ar[i]
# Set training image shapes
shapes = [[1, 1]] * nb
for i in range(nb):
ari = ar[bi == i]
mini, maxi = ari.min(), ari.max()
if maxi < 1:
shapes[i] = [maxi, 1]
elif mini > 1:
shapes[i] = [1, 1 / mini]
self.batch_shapes = np.ceil(np.array(shapes) * img_size / 32.).astype(np.int) * 32
# Preload labels (required for weighted CE training)
self.imgs = [None] * n
self.labels = [None] * n
preload_labels = False
if preload_labels:
self.labels = [np.zeros((0, 5))] * n
iter = tqdm(self.label_files, desc='Reading labels') if n > 10 else self.label_files
extract_bounding_boxes = False
for i, file in enumerate(iter):
try:
with open(file, 'r') as f:
l = np.array([x.split() for x in f.read().splitlines()], dtype=np.float32)
if l.shape[0]:
assert l.shape[1] == 5, '> 5 label columns: %s' % file
assert (l >= 0).all(), 'negative labels: %s' % file
assert (l[:, 1:] <= 1).all(), 'non-normalized or out of bounds coordinate labels: %s' % file
self.labels[i] = l
# Extract object detection boxes for a second stage classifier
if extract_bounding_boxes:
p = Path(self.img_files[i])
img = cv2.imread(str(p))
h, w, _ = img.shape
for j, x in enumerate(l):
f = '%s%sclassification%s%g_%g_%s' % (
p.parent.parent, os.sep, os.sep, x[0], j, p.name)
if not os.path.exists(Path(f).parent):
os.makedirs(Path(f).parent) # make new output folder
box = xywh2xyxy(x[1:].reshape(-1, 4)).ravel()
box = np.clip(box, 0, 1) # clip boxes outside of image
result = cv2.imwrite(f, img[int(box[1] * h):int(box[3] * h),
int(box[0] * w):int(box[2] * w)])
if not result:
print('stop')
except:
pass # print('Warning: missing labels for %s' % self.img_files[i]) # missing label file
assert len(np.concatenate(self.labels, 0)) > 0, 'No labels found. Incorrect label paths provided.'
def __init__(self, path, img_size=416, batch_size=16, augment=False, hyp=None, rect=True, image_weights=False,
cache_images=False):
path = str(Path(path)) # os-agnostic
with open(path, 'r') as f:
self.img_files = [x.replace('/', os.sep) for x in f.read().splitlines() # os-agnostic
if os.path.splitext(x)[-1].lower() in img_formats]
n = len(self.img_files)
bi = np.floor(np.arange(n) / batch_size).astype(np.int) # batch index
nb = bi[-1] + 1 # number of batches
assert n > 0, 'No images found in %s' % path
self.n = n
self.batch = bi # batch index of image
self.img_size = img_size
self.augment = augment
self.hyp = hyp
self.image_weights = image_weights
self.rect = False if image_weights else rect
# Define labels
class_ids = ['n02691156', 'n02419796', 'n02131653', 'n02834778',
'n01503061', 'n02924116', 'n02958343', 'n02402425',
'n02084071', 'n02121808', 'n02503517', 'n02118333',
'n02510455', 'n02342885', 'n02374451', 'n02129165',
'n01674464', 'n02484322', 'n03790512', 'n02324045',
'n02509815', 'n02411705', 'n01726692', 'n02355227',
'n02129604', 'n04468005', 'n01662784', 'n04530566',
'n02062744', 'n02391049']
self.id2idx = {id: idx for idx, id in enumerate(class_ids)}
self.label_files = [x.replace('Data', 'Annotations').replace(os.path.splitext(x)[-1], '.xml')
for x in self.img_files]
# Rectangular Training https://github.com/ultralytics/yolov3/issues/232
if self.rect:
# Read image shapes
sp = 'data' + os.sep + path.replace('.txt', '.shapes').split(os.sep)[-1] # shapefile path
try:
with open(sp, 'r') as f: # read existing shapefile
s = [x.split() for x in f.read().splitlines()]
assert len(s) == n, 'Shapefile out of sync'
except:
s = [exif_size(Image.open(f)) for f in tqdm(self.img_files, desc='Reading image shapes')]
np.savetxt(sp, s, fmt='%g') # overwrites existing (if any)
# Sort by aspect ratio
s = np.array(s, dtype=np.float64)
ar = s[:, 1] / s[:, 0] # aspect ratio
i = ar.argsort()
self.img_files = [self.img_files[i] for i in i]
self.label_files = [self.label_files[i] for i in i]
self.shapes = s[i]
ar = ar[i]
# Set training image shapes
shapes = [[1, 1]] * nb
for i in range(nb):
ari = ar[bi == i]
mini, maxi = ari.min(), ari.max()
if maxi < 1:
shapes[i] = [maxi, 1]
elif mini > 1:
shapes[i] = [1, 1 / mini]
self.batch_shapes = np.ceil(np.array(shapes) * img_size / 32.).astype(np.int) * 32
# Preload labels (required for weighted CE training)
self.imgs = [None] * n
self.labels = [None] * n
if augment or image_weights: # cache labels for faster training
self.labels = [np.zeros((0, 5))] * n
extract_bounding_boxes = False
pbar = tqdm(self.label_files, desc='Reading labels')
nm, nf, ne = 0, 0, 0 # number missing, number found, number empty
for i, file in enumerate(pbar):
try:
boxes = []
root = ET.parse(file).getroot()
im_w = int(root.find('size/width').text)
im_h = int(root.find('size/height').text)
for obj in root.findall('object'):
c = obj.find('name').text
if c not in self.id2idx:
continue
cls = self.id2idx[c]
x1 = int(obj.find('bndbox/xmin').text)
y1 = int(obj.find('bndbox/ymin').text)
x2 = int(obj.find('bndbox/xmax').text)
y2 = int(obj.find('bndbox/ymax').text)
x = 0.5 * (x1 + x2) / im_w
y = 0.5 * (y1 + y2) / im_h
ww = (x2 - x1) / im_w
hh = (y2 - y1) / im_h
boxes.append([cls, x, y, ww, hh])
l = np.array(boxes, dtype=np.float32)
except:
nm += 1 # print('missing labels for image %s' % self.img_files[i]) # file missing
continue
if l.shape[0]:
assert l.shape[1] == 5, '> 5 label columns: %s' % file
assert (l >= 0).all(), 'negative labels: %s' % file
assert (l[:, 1:] <= 1).all(), 'non-normalized or out of bounds coordinate labels: %s' % file
self.labels[i] = l
nf += 1 # file found
# Extract object detection boxes for a second stage classifier
if extract_bounding_boxes:
p = Path(self.img_files[i])
img = cv2.imread(str(p))
h, w, _ = img.shape
for j, x in enumerate(l):
f = '%s%sclassifier%s%g_%g_%s' % (p.parent.parent, os.sep, os.sep, x[0], j, p.name)
if not os.path.exists(Path(f).parent):
os.makedirs(Path(f).parent) # make new output folder
box = xywh2xyxy(x[1:].reshape(-1, 4)).ravel()
b = np.clip(box, 0, 1) # clip boxes outside of image
ret_val = cv2.imwrite(f, img[int(b[1] * h):int(b[3] * h), int(b[0] * w):int(b[2] * w)])
assert ret_val, 'Failure extracting classifier boxes'
else:
ne += 1 # file empty
pbar.desc = 'Reading labels (%g found, %g missing, %g empty for %g images)' % (nf, nm, ne, n)
assert nf > 0, 'No labels found. Recommend correcting image and label paths.'
# Cache images into memory for faster training (~5GB)
if cache_images and augment: # if training
for i in tqdm(range(min(len(self.img_files), 10000)), desc='Reading images'): # max 10k images
img_path = self.img_files[i]
img = cv2.imread(img_path) # BGR
assert img is not None, 'Image Not Found ' + img_path
r = self.img_size / max(img.shape) # size ratio
if self.augment and r < 1: # if training (NOT testing), downsize to inference shape
h, w, _ = img.shape
img = cv2.resize(img, (int(w * r), int(h * r)), interpolation=cv2.INTER_LINEAR) # or INTER_AREA
self.imgs[i] = img
# Detect corrupted images https://medium.com/joelthchao/programmatically-detect-corrupted-image-8c1b2006c3d3
detect_corrupted_images = False
if detect_corrupted_images:
from skimage import io # conda install -c conda-forge scikit-image
for file in tqdm(self.img_files, desc='Detecting corrupted images'):
try:
_ = io.imread(file)
except:
print('Corrupted image detected: %s' % file)