Solar Panel Google Satellite View Image Recognition Scraper

Hello, I need an application built to scrape google satellite images for pictures of solar panels. A good understanding of image recognition is essential, I am expecting that false positives will happen, and this is not a problem, but the accuracy of detection should be as good as possible, without missing potential positives. The application will need to do the following: • Scrape google API for solar panels. Ability to search within X radius of entered location. E.g. 40 miles of London. • Capture solar panel coordinate using google API, the google API is based on latitude and longtitude coordinates, the coordinate shall be taken from approximately the middle of solar panels. • If the google API allows you to do so, it will capture these coordinates in latitude and longtitude AND british national grid (these are 2 types of coordinates). If not, the standard format will do. • Hold collected data in database and do not allow duplicates. • Ability to export whole database or each “new record” each time programme is ran in csv format. • Ability to pause and resume a search. • Feature to save the search/view previous searches. E.g. 40 miles of London = 6578 results, 45% completed, and you can download those results or resume. 100 miles of Edinburgh = 34,000 results, 100% completed, and ability to download those results. This can be a web or desktop based application in any coding language. But it must be robust to run for hours and hold large datasets running into the tens or even hundreds of thousands, because of this I think server based app is needed. I can provide a VPS or a full server if necessary for multi threaded searches.