I'm trying to use the OpenAI's new vision feature where we can send image through Api. What I want is the user to upload an image in the form and send this image to OpenAI. But I can't access this image properly in deluge script. There are also some constraints on the image format when sending to OpenAI. the image should be in base64 format, or URL of the image should be passed.

 So how do I do that?