def tesseract_ocr(imgname, type='PagesWrapper'):
api = tesseract.TessBaseAPI()
api.SetOutputName("outputName")
api.Init(".", "eng", tesseract.OEM_DEFAULT)
api.SetPageSegMode(tesseract.PSM_AUTO)
if type == 'PagesWrapper':
result = tesseract.ProcessPagesWrapper(imgname, api)
elif type == 'PagesFileStream':
result = tesseract.ProcessPagesFileStream(mImgFile, api)
elif type == 'PagesRaw':
result = tesseract.ProcessPagesRaw(mImgFile, api)
elif type == 'PagesBuffer':
mBuffer = open(imgname).read()
result = tesseract.ProcessPagesBuffer(mBuffer, len(mBuffer), api)
return result
# -*- coding: utf-8 -*-
#from __future__ import print_function
import tesseract
import ctypes
import os
#print "HAVE_LIBLEPT=",tesseract.isLibLept()
#print dir("tesseract")
#print tesseract.MAX_NUM_INT_FEATURES
api = tesseract.TessBaseAPI()
api.SetOutputName("outputName")
#api.Init(".","eng")
api.Init(".", "eng", tesseract.OEM_DEFAULT)
api.SetPageSegMode(tesseract.PSM_AUTO)
mImgFile = "eurotext.jpg"
result = tesseract.ProcessPagesWrapper(mImgFile, api)
print "result(ProcessPagesWrapper)=", result
#api.ProcessPages(mImgFile,None, 0, result)
#print "abc"
result = tesseract.ProcessPagesFileStream(mImgFile, api)
print "result(ProcessPagesFileStream)=", result
result = tesseract.ProcessPagesRaw(mImgFile, api)
print "result(ProcessPagesRaw)", result
f = open(mImgFile, "rb")
mBuffer = f.read()
f.close()
result = tesseract.ProcessPagesBuffer(mBuffer, len(mBuffer), api)
mBuffer = None
print "result(ProcessPagesBuffer)=", result
y1 += options.line_gap
if y1 > height:
y1 = height
pg_box = (x0, y0, x1, y1)
region = img.crop(pg_box)
region.save(tmpimgname, "TIFF")
y0 = (height - y1)
else:
x0 = 0
y0 = 0
x1 = 0
y1 = 0
result = ""
if (square.y1 - square.y0) > options.box_threshold:
orig_result = tesseract.ProcessPagesWrapper(tmpimgname, api) + ""
result = orig_result.replace("\n", "")
result = result.replace("\t", "")
result = result.strip()
if len(result) > 0:
# print "RESULTS-------------------------------------->"
# print "%d - Result= %s" % (len(result),result)
# print "<--------------------------------------RESULTS"
if not options.hadoop:
file.write("%s\n" % result)
coordtemp = tempfile.NamedTemporaryFile()
result = tesseract.ExtractResultsWrapper(api, coordtemp.name,
len(orig_result), "")
#print "len", result
im = cam.get_image(surface)
pygame.display.update()
screen.blit(im,(0,0))
#if i==400:
# pygame.image.save(im, "a.jpg")
# img="a.jpg"
# result = tesseract.ProcessPagesWrapper(img,api)
# print result
# print "ok"
for event in pygame.event.get():
# Shutdown with X button
if event.type==pygame.QUIT:
sys.exit()
# Shutdown with ESC
elif event.type == KEYDOWN:
if event.key == K_ESCAPE:
sys.exit()
i=i+1
'''
im = Image.open("5.jpg")
#im=im.rotate(1)
im.save("e.jpg")
im2 = im.convert("L")
im2.save("b.jpg")
threshold = 100
im = im2.point(lambda p: p > threshold and 255)
im.save("d.jpg")
img = "d.jpg"
result = tesseract.ProcessPagesWrapper(img, api)
print result
print "ok"