Good morning Mr.Veser, I hope you are and your family is staying safe. I understand your question is to create 3D representation in a usable format. I am not familiar with Python and TensorFlow. Is this for coding? I do know that I have used my DSLR camera images and drone images together to create a 3D overview of a project but for a different purpose. Have you by any chance looked into any drone software which takes images and creates the 3D representation? I do not know if this would help but might be a different approach. Keep me up to date on your progress, I would be interested to see what you come up with.

Thanks

Mike T.

Hello Mr. Tucker,

thank you very much for your answer, I hope you are doing well. This sounds very interesting, I never thought about drones and related software.

There is really a lot of software for 3D translation of drone images which are useful for generating different file formats, here is one example I stumbled across (https://www.youtube.com/watch?v=8jzxdzcDGMQ).

Yes right, this is for coding in Python. I want to use this data as input for a neural network to generate scenery based on the 3D photos. I will keep ypu updated if I find anything new.

Best regards,

Raffael V.

what you are describing is called Reflectance Transformation Imaging, RTI. if you google it it is easy to find info, technical specifications, different high/low teck versions (for example, RGBN, IRRTI), etc.

if you are photographing something that is threedimentional (for example, a statue) you can use 3d scanning apps that are based on photographs, like 3dzephyr or any photogrammetry software. some of these are free and u can find them in your android/ios application depository

alternatively u can use a "stacking" method in photoshop which will give you a lot of data but not real 3d (take many images with different lighting, PS: file/script/load images to stack-allow photo alignment and then file/script/export layers to file and save file as .psd)