nnunetv2系列：使用默认的预测类推理2D数据

这里参考源代码nnUNet/nnunetv2/inference/predict_from_raw_data.py中给的示例进行调整和测试。

代码示例

from torch import device
from nnunetv2.inference.predict_from_raw_data import nnUNetPredictor

# from nnunetv2.paths import (
#     nnUNet_results,
#     # nnUNet_raw
# )
# from batchgenerators.utilities.file_and_folder_operations import join

from time import time

if __name__ == "__main__":
    start = time()
    # instantiate the nnUNetPredictor
    predictor = nnUNetPredictor(
        tile_step_size=0.5,
        use_gaussian=True,
        use_mirroring=True,
        perform_everything_on_device=True,
        device=device("cuda", 0),
        verbose=False,
        verbose_preprocessing=False,
        allow_tqdm=True,
    )
    # initializes the network architecture, loads the checkpoint
    predictor.initialize_from_trained_model_folder(
        # 直接使用绝对路径，替换join方法
        "/home/bio/family/segmenation/nnUNet/nnUNet_results/Dataset500_ScleraIrisSegmentation/nnUNetTrainer__nnUNetPlans__2d",
        # join(
        #     nnUNet_results,
        #     "Dataset500_ScleraIrisSegmentation/nnUNetTrainer__nnUNetPlans__2d"
        # ),
        use_folds=(0,),
        checkpoint_name="checkpoint_best.pth",
    )

    # variant 1: give input and output folders
    # 使用绝对路径，否则会报错
    # 推荐内部注释生成json文件的代码，否则默认会生成json文件
    predictor.predict_from_files(
         # 实际测试发现，必须先转成nnunet格式，再进行预测，数据名称应该为*_0000.png这类的
        "/home/bio/family/segmenation/nnUNet/afamily_test/inference/imagesTr",
        "/home/bio/family/segmenation/nnUNet/afamily_test/inference/imagesTr_predict",
        save_probabilities=False,
        overwrite=False,
        num_processes_preprocessing=1,
        num_processes_segmentation_export=1,
        folder_with_segs_from_prev_stage=None,
        num_parts=1,
        part_id=0,
    )

    print(f"Time taken: {time() - start}")