google-vision

I want to use the legacy text_detection and document_text_detection model. (refer: https://cloud.google.com/vision/docs/service-announcements)
Im trying it this way using features: