easyocr插件使用

1、上传签字图片（图片背景颜色，和图片的大小会影响文字识别准确率）

2、服务端代码如下

from flask import Flask, request, Response
import easyocr
import json
from hanziconv import HanziConv
reader = easyocr.Reader(['ch_sim', 'en'], model_storage_directory='.EasyOCR/model', download_enabled=False)
app = Flask(__name__)


@app.route('/ocr', methods=['POST'])
def ocr():
    image = request.files['image']
    image.save('temp.png')
    result = reader.readtext('temp.png', detail=0, paragraph=True)
    chinese_text = ''
    for item in result:
        chinese_text += HanziConv.toSimplified(item)
    text = chinese_text.replace(" ", "")
    data = {
        "code": 0,
        "data": text
    }
    print(data)
    json_data = json.dumps(data)
    return Response(json_data, content_type='application/json;charset=utf-8')


if __name__ == '__main__':
    app.run(host='0.0.0.0', port=8084)

3、客户端安卓调用示例项目地址https://gitee.com/firefix-m/aidl-sign-pdf