Download code/visualize.py from TaxonomyProject/SimulationImage: direct link, hf CLI and curl.
- Browser
- Download file 3.13 kB
-
https://huggingface.co/datasets/TaxonomyProject/SimulationImage/resolve/main/code/visualize.py
- Command line
-
hf download hf://datasets/TaxonomyProject/SimulationImage/code/visualize.py
-
curl -L -o visualize.py https://huggingface.co/datasets/TaxonomyProject/SimulationImage/resolve/main/code/visualize.py
3.13 kB
| import sys, os | |
| from pathlib import Path | |
| import numpy as np | |
| from PIL import Image | |
| # ==== 可按需改动的常量 ==== | |
| IMG_ROOT = "automatic_camera_test_demo" | |
| ANN_ROOT = "automatic_camera_test" | |
| TILE_WIDTH = 512 # 每张缩略图统一宽度 | |
| PADDING = 8 # 小图之间的间距(像素) | |
| OUT_PATH = "thumb.png" | |
| def load_rgba(path: Path): | |
| return Image.open(path).convert("RGBA") if path.exists() else None | |
| def resize_w(img: Image.Image, w: int): | |
| if img is None: return None | |
| ow, oh = img.size | |
| if ow == w: return img | |
| nh = int(round(oh * (w / ow))) | |
| return img.resize((w, nh), Image.BICUBIC) | |
| def load_depth_as_img(npy_path: Path, width: int): | |
| if not npy_path.exists(): return None | |
| d = np.load(npy_path) | |
| # 你的深度处理:clip 到 [0, 2048] | |
| np.clip(d, 0.0, 2048.0, out=d) | |
| # 线性映射到 0~255 灰度(2048 对应 255) | |
| d = (d / 2048.0 * 255.0).astype(np.uint8) | |
| img = Image.fromarray(d, mode="L").convert("RGBA") | |
| return resize_w(img, width) | |
| def make_column(rgb, seg, depth, pad=PADDING): | |
| tiles = [x for x in (rgb, seg, depth) if x is not None] | |
| if not tiles: return None | |
| w = max(t.size[0] for t in tiles) | |
| h = sum(t.size[1] for t in tiles) + pad * (len(tiles) - 1) | |
| col = Image.new("RGBA", (w, h), (255, 255, 255, 255)) | |
| y = 0 | |
| for t in tiles: | |
| col.paste(t, (0, y)) | |
| y += t.size[1] + pad | |
| return col | |
| def main(): | |
| if len(sys.argv) < 3: | |
| print("用法: python make_thumbs_simple.py <scene_name> <pose1,pose2,...> [out.png]") | |
| print("示例: python make_thumbs_simple.py Hospital_A 0001,0007,0042 thumbs.png") | |
| sys.exit(1) | |
| scene = sys.argv[1] | |
| poses = [p.strip() for p in sys.argv[2].split(",") if p.strip()] | |
| out_path = Path(sys.argv[3]) if len(sys.argv) >= 4 else Path(OUT_PATH) | |
| columns = [] | |
| for pose in poses: | |
| rgb_path = Path(IMG_ROOT) / scene / pose / "lit.png" | |
| seg_path = Path(ANN_ROOT) / scene / pose / "seg.png" | |
| depth_path = Path(ANN_ROOT) / scene / pose / "depth.npy" | |
| rgb = resize_w(load_rgba(rgb_path), TILE_WIDTH) | |
| seg = resize_w(load_rgba(seg_path), TILE_WIDTH) | |
| depth = load_depth_as_img(depth_path, TILE_WIDTH) | |
| col = make_column(rgb, seg, depth, PADDING) | |
| if col is None: | |
| print(f"[WARN] 跳过 pose {pose}: 三种文件都缺失") | |
| continue | |
| columns.append(col) | |
| if not columns: | |
| print("[ERROR] 没有任何可用列,检查路径是否正确") | |
| sys.exit(2) | |
| total_w = sum(c.size[0] for c in columns) + PADDING * (len(columns) - 1) | |
| max_h = max(c.size[1] for c in columns) | |
| canvas = Image.new("RGBA", (total_w, max_h), (255, 255, 255, 255)) | |
| x = 0 | |
| for c in columns: | |
| canvas.paste(c, (x, 0)) | |
| x += c.size[0] + PADDING | |
| out_path.parent.mkdir(parents=True, exist_ok=True) | |
| canvas.convert("RGB").save(out_path, "PNG") | |
| print(f"[OK] 已保存: {out_path.resolve()}") | |
| if __name__ == "__main__": | |
| main() |