# ์ฐธ์กฐ ultralytics/data/split_dota.py

์ฐธ๊ณ

์ด ํ์ผ์ https://github.com/ultralytics/ ultralytics/blob/main/ ultralytics/data/split_dota .py์์ ํ์ธํ  ์ ์์ต๋๋ค. ๋ฌธ์ ๋ฅผ ๋ฐ๊ฒฌํ๋ฉด ํ ๋ฆฌํ์คํธ (๐ ๏ธ) ๋ฅผ ํตํด ๋ฌธ์ ๋ฅผ ํด๊ฒฐํ๋๋ก ๋์์ฃผ์ธ์. ๊ฐ์ฌํฉ๋๋ค ๐!

## ultralytics.data.split_dota.bbox_iof(polygon1, bbox2, eps=1e-06)

bbox1๊ณผ bbox2 ์ฌ์ด์ IOF๋ฅผ ๊ณ์ฐํฉ๋๋ค.

๋งค๊ฐ๋ณ์:

์ด๋ฆ ์ ํ ์ค๋ช ๊ธฐ๋ณธ๊ฐ
polygon1 ndarray

๋ค๊ฐํ ์ขํ, (n, 8).

ํ์
bbox2 ndarray

๋ฐ์ด๋ฉ ๋ฐ์ค, (n,4).

ํ์
์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 def bbox_iof(polygon1, bbox2, eps=1e-6): """ Calculate iofs between bbox1 and bbox2. Args: polygon1 (np.ndarray): Polygon coordinates, (n, 8). bbox2 (np.ndarray): Bounding boxes, (n ,4). """ polygon1 = polygon1.reshape(-1, 4, 2) lt_point = np.min(polygon1, axis=-2) # left-top rb_point = np.max(polygon1, axis=-2) # right-bottom bbox1 = np.concatenate([lt_point, rb_point], axis=-1) lt = np.maximum(bbox1[:, None, :2], bbox2[..., :2]) rb = np.minimum(bbox1[:, None, 2:], bbox2[..., 2:]) wh = np.clip(rb - lt, 0, np.inf) h_overlaps = wh[..., 0] * wh[..., 1] left, top, right, bottom = (bbox2[..., i] for i in range(4)) polygon2 = np.stack([left, top, right, top, right, bottom, left, bottom], axis=-1).reshape(-1, 4, 2) sg_polys1 = [Polygon(p) for p in polygon1] sg_polys2 = [Polygon(p) for p in polygon2] overlaps = np.zeros(h_overlaps.shape) for p in zip(*np.nonzero(h_overlaps)): overlaps[p] = sg_polys1[p[0]].intersection(sg_polys2[p[-1]]).area unions = np.array([p.area for p in sg_polys1], dtype=np.float32) unions = unions[..., None] unions = np.clip(unions, eps, np.inf) outputs = overlaps / unions if outputs.ndim == 1: outputs = outputs[..., None] return outputs

DOTA ๋ฐ์ดํฐ ์ธํธ๋ฅผ ๋ก๋ํฉ๋๋ค.

๋งค๊ฐ๋ณ์:

์ด๋ฆ ์ ํ ์ค๋ช ๊ธฐ๋ณธ๊ฐ
data_root str

๋ฐ์ดํฐ ๋ฃจํธ.

ํ์
split str

๋ถํ ๋ ๋ฐ์ดํฐ ์ธํธ๋ train ๋๋ val์ผ ์ ์์ต๋๋ค.

'train'
์ฐธ๊ณ

DOTA ๋ฐ์ดํฐ ์ธํธ์ ๋ํด ๊ฐ์ ํ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ์๋๋ค: - data_root - ์ด๋ฏธ์ง - train - val - ๋ ์ด๋ธ - train - val

์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 def load_yolo_dota(data_root, split="train"): """ Load DOTA dataset. Args: data_root (str): Data root. split (str): The split data set, could be train or val. Notes: The directory structure assumed for the DOTA dataset: - data_root - images - train - val - labels - train - val """ assert split in {"train", "val"}, f"Split must be 'train' or 'val', not {split}." im_dir = Path(data_root) / "images" / split assert im_dir.exists(), f"Can't find {im_dir}, please check your data root." im_files = glob(str(Path(data_root) / "images" / split / "*")) lb_files = img2label_paths(im_files) annos = [] for im_file, lb_file in zip(im_files, lb_files): w, h = exif_size(Image.open(im_file)) with open(lb_file) as f: lb = [x.split() for x in f.read().strip().splitlines() if len(x)] lb = np.array(lb, dtype=np.float32) annos.append(dict(ori_size=(h, w), label=lb, filepath=im_file)) return annos

## ultralytics.data.split_dota.get_windows(im_size, crop_sizes=(1024), gaps=(200), im_rate_thr=0.6, eps=0.01)

์ฐฝ ์ขํ๋ฅผ ๊ฐ์ ธ์ต๋๋ค.

๋งค๊ฐ๋ณ์:

์ด๋ฆ ์ ํ ์ค๋ช ๊ธฐ๋ณธ๊ฐ
im_size tuple

์๋ณธ ์ด๋ฏธ์ง ํฌ๊ธฐ, (๊ฐ๋ก, ์ธ๋ก).

ํ์
crop_sizes List(int

์ฐฝ ํฌ๊ธฐ ์๋ฅด๊ธฐ.

(1024)
gaps List(int

์๋ฌผ ๊ฐ ๊ฐ๊ฒฉ.

(200)
im_rate_thr float

์ฐฝ ์์ญ์ ์๊ณ๊ฐ์ ์ด๋ฏธ์ง ์์ญ์ผ๋ก ๋๋ ๊ฐ์๋๋ค.

0.6
eps float

Epsilon value for math operations.

0.01
์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 def get_windows(im_size, crop_sizes=(1024,), gaps=(200,), im_rate_thr=0.6, eps=0.01): """ Get the coordinates of windows. Args: im_size (tuple): Original image size, (h, w). crop_sizes (List(int)): Crop size of windows. gaps (List(int)): Gap between crops. im_rate_thr (float): Threshold of windows areas divided by image ares. eps (float): Epsilon value for math operations. """ h, w = im_size windows = [] for crop_size, gap in zip(crop_sizes, gaps): assert crop_size > gap, f"invalid crop_size gap pair [{crop_size} {gap}]" step = crop_size - gap xn = 1 if w <= crop_size else ceil((w - crop_size) / step + 1) xs = [step * i for i in range(xn)] if len(xs) > 1 and xs[-1] + crop_size > w: xs[-1] = w - crop_size yn = 1 if h <= crop_size else ceil((h - crop_size) / step + 1) ys = [step * i for i in range(yn)] if len(ys) > 1 and ys[-1] + crop_size > h: ys[-1] = h - crop_size start = np.array(list(itertools.product(xs, ys)), dtype=np.int64) stop = start + crop_size windows.append(np.concatenate([start, stop], axis=1)) windows = np.concatenate(windows, axis=0) im_in_wins = windows.copy() im_in_wins[:, 0::2] = np.clip(im_in_wins[:, 0::2], 0, w) im_in_wins[:, 1::2] = np.clip(im_in_wins[:, 1::2], 0, h) im_areas = (im_in_wins[:, 2] - im_in_wins[:, 0]) * (im_in_wins[:, 3] - im_in_wins[:, 1]) win_areas = (windows[:, 2] - windows[:, 0]) * (windows[:, 3] - windows[:, 1]) im_rates = im_areas / win_areas if not (im_rates > im_rate_thr).any(): max_rate = im_rates.max() im_rates[abs(im_rates - max_rate) < eps] = 1 return windows[im_rates > im_rate_thr]

## ultralytics.data.split_dota.get_window_obj(anno, windows, iof_thr=0.7)

๊ฐ ์ฐฝ์ ๋ํ ๊ฐ์ฒด๋ฅผ ๊ฐ์ ธ์ต๋๋ค.

์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 133 134 135 136 137 138 139 140 141 142 143 144 def get_window_obj(anno, windows, iof_thr=0.7): """Get objects for each window.""" h, w = anno["ori_size"] label = anno["label"] if len(label): label[:, 1::2] *= w label[:, 2::2] *= h iofs = bbox_iof(label[:, 1:], windows) # Unnormalized and misaligned coordinates return [(label[iofs[:, i] >= iof_thr]) for i in range(len(windows))] # window_anns else: return [np.zeros((0, 9), dtype=np.float32) for _ in range(len(windows))] # window_anns

## ultralytics.data.split_dota.crop_and_save(anno, windows, window_objs, im_dir, lb_dir)

์ด๋ฏธ์ง๋ฅผ ์๋ฅด๊ณ  ์ ๋ ์ด๋ธ์ ์ ์ฅํฉ๋๋ค.

๋งค๊ฐ๋ณ์:

์ด๋ฆ ์ ํ ์ค๋ช ๊ธฐ๋ณธ๊ฐ
anno dict

๋ค์์ ํฌํจํ ์ฃผ์ ๋์๋๋ฆฌ filepath, label, ori_size ๋ฅผ ํค๋ก ์ฌ์ฉํฉ๋๋ค.

ํ์
windows list

์ฐฝ ์ขํ ๋ชฉ๋ก์๋๋ค.

ํ์
window_objs list

๊ฐ ์ฐฝ ์์ ๋ ์ด๋ธ ๋ชฉ๋ก์๋๋ค.

ํ์
im_dir str

์ด๋ฏธ์ง์ ์ถ๋ ฅ ๋๋ ํฐ๋ฆฌ ๊ฒฝ๋ก์๋๋ค.

ํ์
lb_dir str

๋ ์ด๋ธ์ ์ถ๋ ฅ ๋๋ ํ ๋ฆฌ ๊ฒฝ๋ก์๋๋ค.

ํ์
์ฐธ๊ณ

DOTA ๋ฐ์ดํฐ ์ธํธ์ ๋ํด ๊ฐ์ ํ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ์๋๋ค: - data_root - ์ด๋ฏธ์ง - train - val - ๋ ์ด๋ธ - train - val

์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 def crop_and_save(anno, windows, window_objs, im_dir, lb_dir): """ Crop images and save new labels. Args: anno (dict): Annotation dict, including `filepath`, `label`, `ori_size` as its keys. windows (list): A list of windows coordinates. window_objs (list): A list of labels inside each window. im_dir (str): The output directory path of images. lb_dir (str): The output directory path of labels. Notes: The directory structure assumed for the DOTA dataset: - data_root - images - train - val - labels - train - val """ im = cv2.imread(anno["filepath"]) name = Path(anno["filepath"]).stem for i, window in enumerate(windows): x_start, y_start, x_stop, y_stop = window.tolist() new_name = f"{name}__{x_stop - x_start}__{x_start}___{y_start}" patch_im = im[y_start:y_stop, x_start:x_stop] ph, pw = patch_im.shape[:2] cv2.imwrite(str(Path(im_dir) / f"{new_name}.jpg"), patch_im) label = window_objs[i] if len(label) == 0: continue label[:, 1::2] -= x_start label[:, 2::2] -= y_start label[:, 1::2] /= pw label[:, 2::2] /= ph with open(Path(lb_dir) / f"{new_name}.txt", "w") as f: for lb in label: formatted_coords = ["{:.6g}".format(coord) for coord in lb[1:]] f.write(f"{int(lb[0])} {' '.join(formatted_coords)}\n")

## ultralytics.data.split_dota.split_images_and_labels(data_root, save_dir, split='train', crop_sizes=(1024), gaps=(200))

์ด๋ฏธ์ง์ ๋ ์ด๋ธ์ ๋ชจ๋ ๋ถํ ํฉ๋๋ค.

์ฐธ๊ณ

DOTA ๋ฐ์ดํฐ ์ธํธ์ ๋ํด ๊ฐ์ ํ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ์๋๋ค: - data_root - ์ด๋ฏธ์ง - split - ๋ ์ด๋ธ - split ์ ์ถ๋ ฅ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ๋ - save_dir - ์ด๋ฏธ์ง - split - ๋ ์ด๋ธ - split

์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 def split_images_and_labels(data_root, save_dir, split="train", crop_sizes=(1024,), gaps=(200,)): """ Split both images and labels. Notes: The directory structure assumed for the DOTA dataset: - data_root - images - split - labels - split and the output directory structure is: - save_dir - images - split - labels - split """ im_dir = Path(save_dir) / "images" / split im_dir.mkdir(parents=True, exist_ok=True) lb_dir = Path(save_dir) / "labels" / split lb_dir.mkdir(parents=True, exist_ok=True) annos = load_yolo_dota(data_root, split=split) for anno in tqdm(annos, total=len(annos), desc=split): windows = get_windows(anno["ori_size"], crop_sizes, gaps) window_objs = get_window_obj(anno, windows) crop_and_save(anno, windows, window_objs, str(im_dir), str(lb_dir))

## ultralytics.data.split_dota.split_trainval(data_root, save_dir, crop_size=1024, gap=200, rates=(1.0))

DOTA์ ๋ถํ  ๊ธฐ์ฐจ ๋ฐ ๋ฐธ ์ธํธ.

์ฐธ๊ณ

DOTA ๋ฐ์ดํฐ ์ธํธ์ ๋ํด ๊ฐ์ ํ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ์๋๋ค: - data_root - ์ด๋ฏธ์ง - train - val - ๋ ์ด๋ธ - train - val ๋ก ์ง์ ํ๊ณ  ์ถ๋ ฅ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ๋ ๋ค์๊ณผ ๊ฐ์ต๋๋ค: - save_dir - ์ด๋ฏธ์ง - train - val - ๋ ์ด๋ธ - train - val

์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 def split_trainval(data_root, save_dir, crop_size=1024, gap=200, rates=(1.0,)): """ Split train and val set of DOTA. Notes: The directory structure assumed for the DOTA dataset: - data_root - images - train - val - labels - train - val and the output directory structure is: - save_dir - images - train - val - labels - train - val """ crop_sizes, gaps = [], [] for r in rates: crop_sizes.append(int(crop_size / r)) gaps.append(int(gap / r)) for split in ["train", "val"]: split_images_and_labels(data_root, save_dir, split, crop_sizes, gaps)

## ultralytics.data.split_dota.split_test(data_root, save_dir, crop_size=1024, gap=200, rates=(1.0))

DOTA์ ๋ถํ  ํ์คํธ ์ธํธ, ๋ ์ด๋ธ์ ์ด ์ธํธ์ ํฌํจ๋์ง ์์ต๋๋ค.

์ฐธ๊ณ

DOTA ๋ฐ์ดํฐ ์ธํธ์ ๋ํด ๊ฐ์ ํ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ์๋๋ค: - data_root - images - test ์ด๋ฉฐ ์ถ๋ ฅ ๋๋ ํ ๋ฆฌ ๊ตฌ์กฐ๋ - save_dir - images - test

์ ์์ค ์ฝ๋ ultralytics/data/split_dota.py
 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 def split_test(data_root, save_dir, crop_size=1024, gap=200, rates=(1.0,)): """ Split test set of DOTA, labels are not included within this set. Notes: The directory structure assumed for the DOTA dataset: - data_root - images - test and the output directory structure is: - save_dir - images - test """ crop_sizes, gaps = [], [] for r in rates: crop_sizes.append(int(crop_size / r)) gaps.append(int(gap / r)) save_dir = Path(save_dir) / "images" / "test" save_dir.mkdir(parents=True, exist_ok=True) im_dir = Path(data_root) / "images" / "test" assert im_dir.exists(), f"Can't find {im_dir}, please check your data root." im_files = glob(str(im_dir / "*")) for im_file in tqdm(im_files, total=len(im_files), desc="test"): w, h = exif_size(Image.open(im_file)) windows = get_windows((h, w), crop_sizes=crop_sizes, gaps=gaps) im = cv2.imread(im_file) name = Path(im_file).stem for window in windows: x_start, y_start, x_stop, y_stop = window.tolist() new_name = f"{name}__{x_stop - x_start}__{x_start}___{y_start}" patch_im = im[y_start:y_stop, x_start:x_stop] cv2.imwrite(str(save_dir / f"{new_name}.jpg"), patch_im)

Created 2024-01-05, Updated 2024-06-02
Authors: glenn-jocher (4), Burhan-Q (1)