1.獲取需要OCR識別的圖片

from PIL import Image
import pytesseract

def screenshots_picture(driver,locator):
    '''
    擷取需要被ocr識別的圖片
    :param driver:瀏覽器driver
    :param locator: 元素
    :param fileName: 截圖檔名稱
    :param screenshots_fileName: 識別圖片檔名稱
    :return: 識別碼
    注意：
        # 如果是retina螢幕,必須要加這個不然，就會出現擷取驗證的圖錯誤
        # dpr = driver.execute_script('return window.devicePixelRatio')
        # im = Image.open(picture_name1)
        # img = im.crop((left * dpr, top * dpr, right * dpr, height * dpr))

    
 
'''
    try:
        # 擷取當前網頁，該網頁有我們需要的驗證碼
        name = f'{time.time()}.png'
        fileName = filePictruePath(name)
        driver.save_screenshot(fileName)
        #定位到驗證碼的元素
        imgelement = driver.find_element(*locator)
        # 獲取驗證碼x,y軸座標
        location = imgelement.location
        x 
 
= int(location['x'])
        y = int(location['y'])
        #獲取驗證碼的長寬
        size = imgelement.size
        width = int(size['width'])
        height = int(size['height'])
        dpr = driver.execute_script('return window.devicePixelRatio')
        # 得到要被截圖的位置座標,通過兩點定位要截圖的位置
        rangle = (x*dpr,y*dpr,(x+width)*dpr,(y+height)*dpr)
        
 
#開啟螢幕截圖
        open_fileName = Image.open(fileName)
        # 使用Image的crop函式，從截圖中再次擷取我們需要的區域
        screenshots = open_fileName.crop(rangle)
        #儲存已擷取的驗證碼圖片
        ocr_name = f'{time.time()}ocr.png'
        screenshots_fileName = filePictruePath(ocr_name)
        screenshots.save(screenshots_fileName)

        return screenshots_fileName
    except Exception:
        return None

2.OCR識別圖片

def ocr_code(screenshots_fileName):
    '''
    ocr識別方法
    :param screenshots_fileName: 被識別的檔名稱
    :return: 識別資訊
    '''
    # 開啟儲存的圖片
    open_stream = Image.open(screenshots_fileName)
    # 使用pytesseract中的image_to_string方法獲取識別驗證碼
    identify_text = pytesseract.image_to_string(open_stream).strip()
    print(identify_text)
    # 過濾掉會受影響的符號
    identify_text = filter_str(identify_text)
    return identify_text

用到的方法：

def filePictruePath(name):
    '''
    生成檔案路徑
    :param name:
    :return:
    '''
    file_dir = f"{os.path.dirname(os.path.dirname(__file__))}/screenshot/"
    if os.path.exists(file_dir) and os.path.isdir(file_dir):
        pass
    else:
        os.mkdir(file_dir)
    return os.path.join(file_dir,name)

def filter_str(args):
    '''
    過濾字串中的無效字元
    :param args: 只留數字以及字串
    :return:
    '''
    new_str = str(args)
    new_str = ''.join(new_str.strip().split())
    str_list = []
    for i in new_str:
        if '0' <= i and i <= '9':
            str_list.append(i)
        elif i.upper() >= 'A' and i.upper() <= 'Z':
            str_list.append(i)
    return ''.join(str_list)

會出現下面的錯誤：

具體解決辦法見：https://blog.csdn.net/qq_31362767/article/details/107891185