加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
文件
克隆/下载
ocr.py 661 Bytes
一键复制 编辑 原始数据 按行查看 历史
import os
from cnocr import CnOcr
import re
import sys
def is_image_file(filename):
return any(filename.endswith(ext) for ext in ['.png','.jpg','jpeg','PNG','JPG','JPEG'])
dir=sys.argv[1]+'/'
print(dir)
tem = os.listdir(dir)
jpg = [dir+x for x in tem if is_image_file(x)]
pattern = re.compile(r".(png|jpg|jpeg|PNG|JPG|JPEG)")
log = [pattern.sub('.log',j) for j in jpg]
n=len(jpg)
for i in range(n):
ocr = CnOcr()
res = ocr.ocr(jpg[i])
print("Extracting from: ",jpg[i],"(",i,"/",n,") ...")
with open(log[i],'w') as o:
for j in range(len(res)):
tmp = ''.join(res[j][0]) + '\n'
o.write(tmp)
print('Done!')
马建仓 AI 助手
尝试更多
代码解读
代码找茬
代码优化