Thank you for reaching out to Microsoft Q&A forum!
You can try to preprocess the images before sending them to the OCR service. For example, you can use image processing techniques to remove the symbols or enhance the text in the image to make it more readable for the OCR service.
Try using a feature called Custom Vision, which allows you to train the models to recognize specific objects or text. You can use this feature to train a model to recognize the specific symbols that are causing issues and improve the accuracy of the OCR service.
I hope this helps! Thank you.