Clearing outliers in Python using pandas

I am doing a small project (while trying to educate myself in OCR and data analysis), and I am facing a problem I cannot solve.