ppocr INFO: In PP-OCRv3, rec_image_shape parameter defaults to '3, 48, 320', if you are using recognition model with PP-OCRv2 or an older version, please set --rec_image_shape='3,32,320

[dsamarth08]

Please fill in the information below so that we can solve the problem quickly, Thanks !

Describe the bug While executing 'predict_system.py' I have set the rec_image_shape parameter as rec_image_shape='3,32,320'. But even after specifying the value for the parameter system is still not able to recognize the value

Informations (please complete the following information):

• Inference engine for deployment: CPU (windows)
• Why convert to onnx：ONNX is already implemented so to use the additional functionality
• Paddle2ONNX Version: paddle2onnx==1.0.6
• Email/Wechat/Phone: NA

Screenshots:

Additional context: issue is not only observed for paddle2onnx but also observed for paddle. The command which I have executed is given below:

python -m predict_system.py --use_gpu=False \ --cls_model_dir=./inference/ch_ppocr_mobile_v2.0_cls_infer \ --rec_model_dir=./inference/en_PP-OCRv3_rec_infer \ --det_model_dir=./inference/en_PP-OCRv3_det_infer \ --image_dir=doc/imgs_en/img_12.jpg \ --rec_char_dict_path=ppocr/utils/en_dict.txt \ --rec_image_shape='3,32,320'

(Note: I have also tried giving the shape in different but the issue still persists. --rec_image_shape="3,32,320",
--rec_image_shape "3,32,320", --rec_image_shape 3,32,320

[tsufbechor]

I am facing same issue, have you solved it ?

[mengqiaofa]

I am facing same issue, have you solved it?

[Zheng-Bicheng]

@GreatV

[GreatV]

This is a notice, not an error. For usage details, please refer to: https://paddlepaddle.github.io/PaddleOCR/latest/ppocr/infer_deploy/python_infer.html#_4