在做分割检测的过程中,有时需要数据清洗并批量提取清洗后数据的标签。
# 导入必要的库
import pandas as pd
import os
import numpy as np
import csv
import shutil
# 数据、标签文件读取
# original.csv:原数据标签文件,包括了所有样本的标签
# new.csv:新的空的标签文件
# file_name:存放数据的文件路径
file_read = open("original.csv", "r")
reader = csv.reader(file_read)
fileheader = ["ImageId", "EncondedPixels"]
file_write = open("new.csv", "w")
writer = csv.writer(file_write)
writer.writerow(fileheader)
big_ship_dic = {}
a = 1
big_ship_set = os.listdir('file_name')
# 循环读取写入
for i in reader:
for item in big_ship_set:
if reader.line_num == 1:
continue
else:
if i[0] == item:
writer.writerow(i)
big_ship