人工智能课程作业-图片文字识别

在conda的虚拟环境，用python随手写的代码，记录学习

识别图片

 path="./"
for path,dir,file in os.walk(path):
    break       
pic=[]
for i in file:
    if(".jpg" in i):
        pic.append(i)

 path指当前目录下，通过walk来遍历当前目录下的文件，寻找出文件后缀名为jpg的文件并存入pic中

申请百度api

from aip import AipOcr
APP_ID = '#####'
API_KEY = '####'
SECRET_KEY = '###'
client = AipOcr(APP_ID, API_KEY, SECRET_KEY)

APP_ID,API_KEY,SECRET_KEY三个都是从百度AI创建应用中得到的，输入即可申请调用对应的api

文字识别

def i2t(picname):
    with open(picname,'rb') as fp:
        image=fp.read()
    dic_result=client.basicAccurate(image)
    res=dic_result['words_result']
    result=''
    for m in res:
        result+=str(m['words'])
 
    fp2=open('./text/a.txt','a')
 
    
    fp2.write(result)
    fp2.close()

basicAccurate表示百度文字识别的精准识别，通过这个函数来实现文字识别功能；把图片读取到image中，用api完成识别，最后把识别内容写入text文件中的a.txt。

附上完整代码

import os
from aip import AipOcr
APP_ID = '28641495'
API_KEY = 'Fsv5EiwDqk3dpe5DG5x3zFyr'
SECRET_KEY = 'wEtWGygI82wWuDBj0LkxGLDxrGN1UyWG'
client = AipOcr(APP_ID, API_KEY, SECRET_KEY)
 
 
path="./"
for path,dir,file in os.walk(path):
	break		
pic=[]
for i in file:
	if(".jpg" in i):
		pic.append(i)
 
 
def i2t(picname):
	with open(picname,'rb') as fp:
		image=fp.read()
	dic_result=client.basicAccurate(image)
	res=dic_result['words_result']
	result=''
	for m in res:
		result+=str(m['words'])
 
	fp2=open('./text/a.txt','a')
 
	
	fp2.write(result)
	fp2.close()
 
 
n=1
for j in pic:
	print("picture%d %s...\n" %(n,j))
	i2t(j)
	n=n+1