图形验证码OCR识别

图形验证码OCR识别。直接上代码…

# _*_coding: UTF-8 _*_ 

import os
import cv2
import pytesseract
from PIL import Image

class OCRTools:
    # 如果没有将tesseract的安装目录添加到系统环境变量中，则需要指定安装路径
    pytesseract.pytesseract.tesseract_cmd = r"D:\Work\tesseract-ocr\tesseract.exe"

    def pic_ocr(self, pic_path):
        # 读取图像
        image = cv2.imread(pic_path)

        # 转换为灰度图像
        gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

        # 中值滤波去噪
        denoised = cv2.medianBlur(gray, 3)

        # 二值化处理
        _, binary_image = cv2.threshold(denoised, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)

        # 反转颜色 验证码底色为深色时，不建议反转
        binary_image = cv2.bitwise_not(binary_image)

        # 腐蚀和膨胀操作，分离连接在一起的字符
        kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (2, 2))
        eroded = cv2.erode(binary_image, kernel, iterations=1)
        dilated = cv2.dilate(eroded, kernel, iterations=1)

        # 轮廓检测
        contours, _ = cv2.findContours(dilated, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

        # 用于存储分割后的字符
        character_images = []

        # 遍历所有检测到的轮廓
        for contour in contours:
            (x, y, w, h) = cv2.boundingRect(contour)
            if w > 5 and h > 10:  # 排除过小的噪声
                char_img = binary_image[y:y+h, x:x+w]
                character_images.append((x, char_img))

        # 对每个字符图像进行识别
        recognized_text = ""
        for x, char_img in sorted(character_images, key=lambda item: item[0]):
            # 放大字符图像
            char_img = cv2.resize(char_img, (char_img.shape[1] * 2, char_img.shape[0] * 2))
            # 转换为 PIL 图像
            char_pil = Image.fromarray(char_img)
            # 使用 Tesseract 进行识别
            char_text = pytesseract.image_to_string(char_pil, config='--psm 10 -c tessedit_char_whitelist=abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789')
            recognized_text += char_text.strip()

        print("识别的文字为：")
        print(recognized_text)
        return recognized_text

if __name__ == '__main__':
	# 验证码路径
    path = os.path.dirname(os.path.dirname(__file__))+os.sep+'image'
    O = OCRTools()
    result = O.pic_ocr(path )