Having trouble doing complex Table Recognition and Extraction from image to excel

[sameedaqds257]

So I have spent a week on it. I know paddle ocr is probably the best choice for it but im not able to figure out how.
Actually I want to extract table from an image and convert to Excel with OCR and proper table structure recognition and recovery.
The data is of an antigen report(medical blood related data), so it included alpha numeric and symbols like + too.
It is also not very crystal clear, although its not blur, but the resolution is around 400*800.
I tried using the predict_table.py file, the Paddle Structure class too but doesn't seem to work out.
The output in excel seems to be some random html tags, when I use table_dict.txt and gives list index out of range error when using en_dict.txt.
Please guide me how can I solve my issue.