folder.py

import os
import tesseract
import cv2.cv as cv
import numpy as np
import locale
import shutil
import uuid
def ocr():
 i=0 
 dst=''
 string=''
 percents=[]
 path="/home/jera/segmentos/"
 dirs = os.listdir(path)
 dirs.sort()

 ### tesseract OCR api setup
 locale.setlocale(locale.LC_NUMERIC,"C")
 #locale.setlocale(locale.LC_ALL, "C")
 api = tesseract.TessBaseAPI()
 api.Init("/usr/share/tesseract-ocr","lus",tesseract.OEM_DEFAULT)
 api.SetVariable("tessedit_char_whitelist", "ABCDEFGHJKLMNOPRSTUVWXYZ")
 api.SetPageSegMode(tesseract.PSM_SINGLE_CHAR)


 #walk through dir 
 for x in dirs:
  # path to individual file
  x=path+x

  #open image
  image=cv.LoadImage((x), cv.CV_LOAD_IMAGE_GRAYSCALE)
  tesseract.SetCvImage(image,api)
  text=api.GetUTF8Text()
  conf=api.MeanTextConf()
  image=None
  if (conf)>76 :
   string+=text[0]
   #dst='/home/jera/prueba6/'+ text[0]+'-'+ str(uuid.uuid1())+'.jpg'
   #shutil.copyfile(x,dst)
   #print string
   percents.append(conf)
   i+=1
   if i==3:
    api.SetVariable("tessedit_char_whitelist", "0123456789")
   if i>=6: 
    os.remove(x)
    break
  #shutil.copyfile(x,dst)
  os.remove(x)

 perc=sorted(percents)
 #Release memory
 api.End()
  
 if len(string)==6:
  return string,perc[0]
 else:
  return None, -1