[V1.2.3] Action dedicated to OCR and Video analysis

[zfoong]

Issue:
Currently, we do have image view action, but it does not producce good OCR result.
Therefore, we need to add more actions for OCT and Video analysis.

Solution:
Create new functions under internal_action_interface.
Create perform_OCR and understand_video actions.