Pytesseract 不接受 pyautogui 屏幕截图、Windows、Python 3.6

Question

What I'm trying to do is to make a screenshot of a number with pyautogui and tranform the number to a string with pytesseract. The code: import pyautogui import time import PIL from PIL import Image import pytesseract

pytesseract.pytesseract.tesseract_cmd = 'C://Program Files (x86)//Tesseract-OCR//tesseract'

# Create image
time.sleep(5)
image = pyautogui.screenshot('projects/output.png', region=(1608, 314, 57, 41))

# Resize image
basewidth = 2000
img = Image.open('projects/output.png')
wpercent = (basewidth/float(img.size[0]))
hsize = int((float(img.size[1])*float(wpercent)))
img = img.resize((basewidth,hsize), PIL.Image.ANTIALIAS)
img.save('projects/output.png')

col = Image.open('projects/output.png')
gray = col.convert('L')
bw = gray.point(lambda x: 0 if x<128 else 255, '1')
bw.save('projects/output.png')

# Image to string
screen = Image.open('projects/output.png')

print(pytesseract.image_to_string(screen, config='tessedit_char_whitelist=0123456789'))

Now it seems that pytesseract doesn't accept the screenshot pyautogui creates. The code runs fine without problems but prints an empty string. If I create an image in paint however, and save it as 'output.png' to the correct folder exactly like the screenshot otherwise made, it does work.

Image output after resize and adjustments

Anyone has an idea where I'm missing something?

Answer 1

Modify the path and try the following:

import numpy as np
from numpy import *
from PIL import Image
from PIL import *
import pytesseract
import cv2


src_path = "C:\\Users\\USERNAME\\Documents\\OCR\\"

def get_region(box):
    #Grabs the region of the box coordinates
    im = ImageGrab.grab(box)
    #Change size of image to 200% of the original size
    a, b, c, d = box
    doubleX = (c - a) * 2
    doubleY = (d - b) * 2
    im.resize((doubleX, doubleY)).save(os.getcwd() + "\\test.png", 'PNG')

def get_string(img_path):
    # Read image with opencv
    img = cv2.imread(img_path)
    # Convert to gray
    img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # Apply dilation and erosion to remove some noise
    kernel = np.ones((1, 1), np.uint8)
    img = cv2.dilate(img, kernel, iterations=1)
    img = cv2.erode(img, kernel, iterations=1)
    # Write image after removed noise
    cv2.imwrite(src_path + "removed_noise.png", img)
    #  Apply threshold to get image with only black and white
    #img = cv2.adaptiveThreshold(img, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 31, 2)
    # Write the image after apply opencv to do some ...
    cv2.imwrite(src_path + "thres.png", img)
    # Recognize text with tesseract for python

    result = pytesseract.image_to_string(Image.open(src_path + "thres.png"))

    return result

def main():
    #Grab the region of the screenshot (box area)
    region = (1354,630,1433,648)
    get_region(region)

    #Output results
    print ("OCR Output: ")
    print (get_string(src_path + "test.png"))

Answer 2

Convert it to a numpy array, pytesseract accepts those.

import numpy as np
import pyautogui

img=np.array(pyautogui.screenshot())
print(pytesseract.image_to_string(screen, config='tessedit_char_whitelist=0123456789'))

Alternatively I would recommend 'mss' for screenshots as they are much faster.

import mss
with mss.mss() as sct:
    img = np.array(sct.grab(sct.monitors[1]))
print(pytesseract.image_to_string(screen, config='tessedit_char_whitelist=0123456789'))