使用python 识别图像文字，快速读取数据

1. 安装opencv

先下载opencv官网的wle文件，然后在该目录，powershell敲击如下代码后tab键，然后安装
pip install opencv-python

3. 安装tesseract 使用

先下载windows版本的，然后安装，使用中文字库的话需要去专门找一下，自带的下不下来，安装完后pip再安装一遍
pip install pytesseract

本地摄像头

import cv2
import numpy as np
import pytesseract as pt
from PIL import Image
path = r"D:\software\tesseract\tesseract.exe"""
pt.tesseract_cmd = path
font = cv2.FONT_HERSHEY_SIMPLEX
cap =cv2.VideoCapture(0,cv2.CAP_DSHOW)
kernel = np.ones((5, 5), np.uint8)
kernel2 = np.ones((7, 7), np.uint8)
while True:
    ret,frame = cap.read()
    frame = frame[0:417, 350:470]
    x,y=frame.shape[0:2]
    frame = cv2.resize(frame,(int(y/2),int(x/2)))
    frame = cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY)
    frame = cv2.dilate(frame, kernel, 1)
    frame = cv2.erode(frame, kernel, 1)
    frame = cv2.erode(frame, kernel2, 1)
    frame = cv2.dilate(frame, kernel, 1)
    cv2.blur(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB), (5, 5))
    ret,thresh1=cv2.threshold(frame,0,255,cv2.THRESH_BINARY | cv2.THRESH_OTSU)    
  #  frame = cv2.erode(frame, kernel, iterations=1)
    text =pt.image_to_string(frame,config='--psm 10 --oem 1 -c tessedit_char_whitelist=0123456789')
    cv2.putText(thresh1, text, (20, 20), font, 1.2, (0, 0, 0), 2)
    cv2.imshow('frame',thresh1)
    c = cv2.waitKey(1)
    if c == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

    

网络ip摄像头（安卓手机使用IP摄像头这个软件，把自己变成电脑摄像头）

#coding=utf-8
import cv2 as cv
import threading
import time
import numpy as np
import pytesseract as pt
from PIL import Image
from queue import Queue
path = r"D:\software\tesseract\tesseract.exe"""
pt.tesseract_cmd = path
font = cv.FONT_HERSHEY_SIMPLEX
kernel = np.ones((5, 5), np.uint8)
kernel2 = np.ones((7, 7), np.uint8)
def show_cam(): 
    while True:  
        image = cv.cvtColor(image_np,cv.COLOR_BGR2GRAY)
        ret,image=cv.threshold(image,0,255,cv.THRESH_BINARY | cv.THRESH_OTSU) 
        text =pt.image_to_string(image,lang="chi_sim")
        q.put(text)
        if cv.waitKey(25) & 0xFF == ord('l'):
            break
    return  

if __name__ == '__main__':
    q = Queue(maxsize=1)
    cv.namedWindow("camera",1)
    #开启ip摄像头
    video="http://admin:admin@192.168.3.3:8081/"
    cap =cv.VideoCapture(video)
    ret, image_np = cap.read()
    thread=threading.Thread(target=show_cam)  
    thread.start()
    i=0
    while True:
        # Start Camera, while true, camera will run
    
        ret, image_np = cap.read()
        
        # Set height and width of webcam
        height = 600
        width = 1000
        if not q.empty():
            text=q.get()
            i=20
        if i>0:
            cv.putText(image_np, text, (100, 100), font, 5, (0, 0, 0), 3)
            i=i-1
        # Set camera resolution and create a break function by pressing 'q'
        cv.imshow('object detection', cv.resize(image_np, (width, height)))
        if cv.waitKey(25) & 0xFF == ord('q'):
            cap.release()
            cv.destroyAllWindows()
            break

然后就可以识别啦！坑比较多，大家慢慢踩！