0

I am trying to detect Date and time from the below image. Both easyOCR and Tessaract are not detecting correctly. I need to detect 2023/05/16 06:40:50 like in below image but easyOCR is detecting it as 2023/057716 06:40850 and Tessaract detecting it as 2623/05/16 66:46:58.

Please click here for the image I am using

I tried with below program

import pytesseract
import cv2
import easyocr

class Test:
    def __init__(self):
        self.text = ''
        self.reader = easyocr.Reader(['en'], gpu=True)
    def extract_time_easyOCR(self):
            print("Inside easyOCR")
            img = cv2.imread("videos\Cropped_Image.jpg")
            result = self.reader.readtext(img)
            for (bbox, text, prob) in result:
                self.text+=text + ' '
            **print("resultfromEasyOCR",self.text)**
            return self.text
    def extract_time_Tessaract(self):
            print("Inside Tessaract")
            image = cv2.imread("videos\Cropped_Image.jpg")
            image_RGB = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
            print("resultfromTessaract",pytesseract.image_to_string(image_RGB,config="--psm 6"))


easyOCRTEST = Test()

easyOCRTEST.extract_time_easyOCR()

easyOCRTEST.extract_time_Tessaract()

and the **output is

Inside easyOCR

resultfromEasyOCR cropped 2023/057716 06:40850

Inside Tessaract

resultfromTessaract 2623/05/16 66:46:58

Can someone please help me to detect correctly? Can I train own model to detect correctly in my requirement or please let me know if there is any other way to detect correctly?

toyota Supra
  • 3,181
  • 4
  • 15
  • 19
Sukumar Reddy
  • 3
  • 3
  • IDK if its even possible to use REGEX in a situation like this, but if you can, it detects character patterns so maybe you could use it to convert some string patterns into the correct ones. – easleyfixed Jun 20 '23 at 18:46

0 Answers0