利用csv文件信息，将图片名信息保存到csv文件当中

本文最后更新于：2023年4月7日下午

我们可以利用train.csv文件信息，再结合给定的文件路径（path）信息，可以将给定字目录下的图片名信息整合到scv文件当中。

train.csv文件格式：

图片名信息：

代码如下:

from glob import glob
import pandas as pd
import os

def enrich_data(df, sdir="train"):
    imgs = glob(os.path.join(DATASET_FOLDER, sdir, "case*", "case*_day*", "scans", "*.png"))
    img_folders = [os.path.dirname(p).split(os.path.sep) for p in imgs]
    img_names = [os.path.splitext(os.path.basename(p))[0].split("_") for p in imgs]
    img_keys = [f"{f[-2]}_slice_{n[1]}" for f, n in zip(img_folders, img_names)]

    # print(img_keys[:5])
    df["img_path"] = df["id"].map({k: p for k, p in zip(img_keys, imgs)})
    df["Case_Day"] = df["id"].map({k: f[-2] for k, f in zip(img_keys, img_folders)})
    df["Case"] = df["id"].apply(lambda x: int(x.split("_")[0].replace("case", "")))
    df["Day"] = df["id"].apply(lambda x: int(x.split("_")[1].replace("day", "")))
    df["Slice"] = df["id"].map({k: int(n[1]) for k, n in zip(img_keys, img_names)})
    df["width"] = df["id"].map({k: int(n[2]) for k, n in zip(img_keys, img_names)})
    df["height"] = df["id"].map({k: int(n[3]) for k, n in zip(img_keys, img_names)})
    df["spacing1"] = df["id"].map({k: float(n[4]) for k, n in zip(img_keys, img_names)})
    df["spacing2"] = df["id"].map({k: float(n[5]) for k, n in zip(img_keys, img_names)})

if __name__ == "__main__":
    # df_ssub = pd.read_csv(os.path.join(DATASET_FOLDER, "sample_submission.csv"))
    DATASET_FOLDER = "D:\compation\kaggle"
    df_ssub = pd.read_csv(os.path.join(DATASET_FOLDER, "train.csv","traines.csv"))
    enrich_data(df_ssub,"traines")
    df_ssub.to_csv("df.csv")
    print(df_ssub["Case_Day"][4])