Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update folder_data_set_loader.py #1

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
36 changes: 36 additions & 0 deletions core/linnaeus/core/loaders/folder_data_set_loader.py
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,42 @@

from linnaeus.core.data_augmentation import preprocessing

class FolderDataSetLoaderYoloFormat(Dataset):
def __init__(self):
"""
self-defined dataset for augmented dataset in yolo format
"""
super(FolderDataSetLoaderYoloFormat, self).__init__()
# root dir of images
self.root_images = "./Data/"
# root dir of annotations
self.root_labels = "./Annotations/"
# obtain the names of labels
self.labels = os.listdir(self.root_labels)

def __getitem__(self, index):
"""
according to index obtain image and its label
:param index:
:return:
"""
image, bbox, category = get_voc_label("/home/wzl/VOC/VOC2007/VOCdevkit/VOC2007/Annotations/" + self.labels[index])
# obtain template, search
template, search, bbox, mapping = transform(image, bbox)
bbox = [(bbox[0] + bbox[2]) / 2, (bbox[1] + bbox[3]) / 2, bbox[2] - bbox[0], bbox[3] - bbox[1]]
# process template and search
torch_template = torch.from_numpy(np.transpose(cv2.cvtColor(template, cv2.COLOR_BGR2RGB), (2, 0, 1)))
torch_search = torch.from_numpy(np.transpose(cv2.cvtColor(search, cv2.COLOR_BGR2RGB), (2, 0, 1)))
# color jitter
torch_template = self.color_jitter(torch_template.unsqueeze(0)).squeeze(0).type(torch.FloatTensor) / 255
torch_search = self.color_jitter(torch_search.unsqueeze(0)).squeeze(0).type(torch.FloatTensor) / 255
return torch_template, torch_search, torch.tensor(bbox) / 255, torch.tensor(mapping), torch.from_numpy(
np.transpose(cv2.resize(image, (256, 256)), (2, 0, 1))), category

def __len__(self):
return len(self.labels)


class FolderDataSetLoader(Dataset):

def __init__(self, path, classes):
Expand Down