def train_image_gen(train_list, batch_size=4, image_size=[1024, 384], crop_offset=690):
# Arrange all indexes
all_batches_index = np.arange(0, len(train_list))
out_images = []
out_masks = []
image_dir = np.array(train_list['image'])
label_dir = np.array(train_list['label'])
while True:
# Random shuffle indexes every epoch
np.random.shuffle(all_batches_index)
for index in all_batches_index:
if os.path.exists(image_dir[index]):
ori_image = cv2.imread(image_dir[index])
ori_mask = cv2.imread(label_dir[index], cv2.IMREAD_GRAYSCALE)
# Crop the top part of the image
# Resize to train size
train_img, train_mask = crop_resize_data(ori_image, ori_mask, image_size, crop_offset)
# Encode
train_mask = encode_labels(train_mask)
# verify_labels(train_mask)
out_images.append(train_img)
out_masks.append(train_mask)
if len(out_images) >= batch_size:
out_images = np.array(out_images)
out_masks = np.array(out_masks)
out_images = out_images[:, :, :, ::-1].transpose(0, 3, 1, 2).astype(np.float32) / (255.0 / 2) - 1
out_masks = out_masks.astype(np.int64)
yield out_images, out_masks
out_images, out_masks = [], []
else:
print(image_dir, 'does not exist.')
def val_image_gen(val_list,
batch_size=4,
image_size=[1024, 384],
crop_offset=690):
all_batches_index = np.arange(0, len(val_list))
out_images = []
out_masks = []
image_dir = np.array(val_list['image'])
label_dir = np.array(val_list['label'])
while True:
np.random.shuffle(all_batches_index)
for index in all_batches_index:
if os.path.exists(image_dir[index]):
ori_image = cv2.imread(image_dir[index])
ori_mask = cv2.imread(label_dir[index], cv2.IMREAD_GRAYSCALE)
val_img, val_mask = crop_val_resize_data(
ori_image, ori_mask, image_size, crop_offset)
val_mask = encode_labels(val_mask)
out_images.append(val_img)
out_masks.append(val_mask)
if len(out_images) >= batch_size:
out_images = np.array(out_images)
out_masks = np.array(out_masks)
out_images = out_images[:, :, :, ::-1].transpose(
0, 3, 1, 2).astype(np.float32) / (255.0 / 2) - 1
out_masks = out_masks.astype(np.int64)
yield out_images, out_masks
out_images, out_masks = [], []
else:
print(image_dir, 'does not exist.')
def __getitem__(self, idx):
ori_image = cv2.imread(self.images[idx])
ori_mask = cv2.imread(self.labels[idx], cv2.IMREAD_GRAYSCALE)
train_img, train_mask = crop_resize_data(ori_image, ori_mask)
train_mask = encode_labels(train_mask)
sample = [train_img.copy(), train_mask.copy()]
if self.transform:
sample = self.transform(sample)
return sample
def __getitem__(self, idx):
img = self.data_list[idx]
label = self.label_list[idx]
img = cv2.imread(img)
label = cv2.imread(label,0)
img,label = crop_resize_data(img,label,self.crop_size,self.offset )
img = tfs.ToTensor()(img)
label = encode_labels(label)
olabel = onehot_label(label)
return img,olabel,label
def get_lane_image(self, idx):
# self.img
# self.label_img 语义分割标签
# self.ins_img 嵌入空间，每个类一个平面
ln = self.labels[idx]
img_name = self.images[idx]
self.img = cv2.imread(img_name, 1) #uint8
self.label_img = cv2.imread(ln, 0) #uint8
# self.label_img = np.array(self.label_img, dtype=np.int32)
self.label_img = encode_labels(self.label_img)
def main(label_paths):
out_array = AverageMeter()
area_dict = {}
for i in range(8):
area_dict[i] = 0
# num_dict = {}
for s, path in enumerate(tqdm(label_paths)):
ori_mask = cv2.imread(path, cv2.IMREAD_GRAYSCALE)
encode_mask = encode_labels(ori_mask).astype('uint8')
get_one_hot = np.eye(8)[encode_mask]
get_one_hot = np.sum(np.sum(get_one_hot, 0), 0)
out_array.update(get_one_hot.copy())
return out_array.avg/100
def __getitem__(self, idx):
# log_file = open("logs/debug.txt", 'a+')
# # print("\nimage: {}\nlabel: {}".format(self.images[idx], self.labels[idx]))
# log_file.writelines("\nimage: {}\nlabel: {}".format(self.images[idx], self.labels[idx]))
# log_file.close()
ori_image = cv2.imread(self.images[idx])
ori_mask = cv2.imread(self.labels[idx], cv2.IMREAD_GRAYSCALE)
train_img, train_mask = crop_resize_data(ori_image, ori_mask)
# Encode
train_mask = encode_labels(train_mask)
sample = [train_img.copy(), train_mask.copy()]
if self.transform:
sample = self.transform(sample)
return sample
def __getitem__(self, idx):
pd.DataFrame(self.images).to_csv(
'C:/Users/Yan/Desktop/项目1/car_lane_detection/logs/image.csv',
index=True)
print(idx)
ori_image = cv2.imread(self.images[idx])
ori_mask = cv2.imread(self.labels[idx], cv2.IMREAD_GRAYSCALE)
train_img, train_mask = crop_resize_data(ori_image, ori_mask)
# Encode
train_mask = encode_labels(train_mask)
sample = [train_img.copy(), train_mask.copy()]
if self.transform:
sample = self.transform(sample)
return sample
def __getitem__(self, idx):
ori_image = cv2.imread(self.images[idx])
if (ori_image is None):
print("idx:{},path:{}".format(idx, self.images[idx]))
ori_mask = cv2.imread(self.labels[idx], cv2.IMREAD_GRAYSCALE)
if (ori_mask is None):
print("idx:{},path:{}".format(idx, self.labels[idx]))
train_img, train_mask = crop_resize_data(ori_image, ori_mask)
# Encode
train_mask = encode_labels(train_mask)
sample = [train_img.copy(), train_mask.copy()]
if self.transform:
sample = self.transform(sample)
return sample
def __getitem__(self, idx):
image_bytes = self.txn.get(self.images[idx].encode())
image_bytes = np.array(bytearray(image_bytes), dtype=np.uint8)
mask_bytes = self.txn.get(self.labels[idx].encode())
mask_bytes = np.array(bytearray(mask_bytes), dtype=np.uint8)
train_img = cv2.imdecode(image_bytes, cv2.IMREAD_COLOR)
train_mask = cv2.imdecode(mask_bytes, cv2.IMREAD_GRAYSCALE)
train_img, train_mask = crop_resize_data(train_img, train_mask)
# Encode
train_mask = encode_labels(train_mask)
sample = [train_img.copy(), train_mask.copy()]
if self.transform:
sample = self.transform(sample)
return sample
def __getitem__(self, idx):
# __getitem__函数，单张读取图片
# 输入：图片的idx， 输出：经过裁剪、放缩、数据增强以及对label转换id的train_image,train_mask
# 按照地址读入ori_image ori_label
ori_image = cv2.imread(self.images[idx])
if (ori_image is None):
print("idx:{},path:{}".format(idx, self.images[idx]))
ori_mask = cv2.imread(self.labels[idx], cv2.IMREAD_GRAYSCALE)
if (ori_mask is None):
print("idx:{},path:{}".format(idx, self.labels[idx]))
# 对ori_image和ori_mask都进行裁剪和放缩得到理想size的图片,train_img,train_mask
train_img, train_mask = crop_resize_data(ori_image, ori_mask)
# 对mask进行id 转换--->train_mask
train_mask = encode_labels(train_mask)
# 复制train_img,train_mask 放入sample列表
sample = [train_img.copy(), train_mask.copy()]
# 有数据增强就做增强
if self.transform:
sample = self.transform(sample)
return sample
def __getitem__(self, idx):
# image 数据读入
ori_image = cv2.imread(self.images[idx])
if ori_image is None:
print('NoneTpye Object ori_image')
time.sleep(60)
ori_image = cv2.imread(self.images[idx])
# label 数据读入
ori_mask = cv2.imread(self.labels[idx], cv2.IMREAD_GRAYSCALE)
if ori_mask is None:
print('NoneTpye Object ori_mask')
time.sleep(60)
ori_mask = cv2.imread(self.labels[idx], cv2.IMREAD_GRAYSCALE)
train_img, train_mask = crop_resize_data(ori_image, ori_mask)
# Encode
train_mask = encode_labels(train_mask)
sample = [train_img.copy(), train_mask.copy()]
if self.transform:
sample = self.transform(sample)
return sample
