Software-Grab screenshots from a video, enter into Google LENS for identification, and extract images and text results

The idea is to take a 1 hour or longer video, extract the individual cars/vehicles in it (will appear in the same orientation) but not necessarily in the same frequency. Once these images are extracted, either via google's API or via automated bot entry, search results need to be extracted e.g Hyundai I10 2012. The video images need to be placed next to the google image and result.