基于paddleocr识别并提取图片中文本信息

1、安装

pip install paddleocr

2、完整的代码

from paddleocr import PaddleOCR, draw_ocr
import numpy as np
import cv2

image_path = "testpic.jpg"

ocr = PaddleOCR(use_angle_cls=True, lang="ch")
result = ocr.ocr(image_path, cls=True)
for line in result:
    print(line)

from PIL import Image

image = Image.open(image_path).convert("RGB")
_result = result[0]
boxes = [item[0] for item in _result]  # 提取文本框坐标
txts = [item[1][0] for item in _result]  # 提取识别的文本
scores = [item[1][1] for item in _result]  # 提取置信度分数

im_show = draw_ocr(image, boxes, txts, scores)
im_show = Image.fromarray(im_show)
im_show.save("result.jpg")
cv2.imshow("resultimg", cv2.cvtColor(np.asarray(im_show), cv2.COLOR_RGB2BGR))
cv2.waitKey(0)
cv2.destroyAllWindows()

3、识别效果