PaddleHubOcr chinese_ocr_db_crnn_mobile预测一张图片需要时间很久,有什么办法解决吗?result=ocr.recognize_text(images=[cv2.imread('/PATH/TO/IMAGE')], text_thresh=0.5) 这行代码的执行时间需要很久
Created by: WJinhai
每次运行的时候都有这两行,能否放在程序启动的时候,这样不需要每次运行都去加载模型 [2020-06-28 15:27:07,018] [ INFO] - Installing chinese_text_detection_db_server module [2020-06-28 15:27:07,019] [ INFO] - Module chinese_text_detection_db_server already installed in /home/image/.paddlehub/modules/chinese_text_detection_db_server