I am in pursuit of the perfect 3dModel , I am building a manual capture flight to give me the granular control of my uneven and jagged terrain
I am trying to understand some best practices to achieve the highest quality up close model of a sculpture of human size sitting on a much larger 3D terrain. I have achieved the results i need for the terrain but not for the up close sculpture
Does Pix4D allow processing of RAW DNG files and if so does the it recognise the information on the file for modelling purposes or the RAW files should be optimised for best visual presentation first?
At what pont do image files SNG , JPG, PNG not contribute to the model quality or is the bigger the image and the closest always improving the model ( ie i am able to fly 1m away from my sculpture at 99% overlap shooting massive RAW files but does that stop helping at some point?)
Is it always best to combine both NADIR top down and oblique angle shots for reconstruction if pursuit of model perfection approximating reality is my goal and are there any other recommendations of best practices for this goal in either capture or processing. ( I can scale up my Cloud GPUs as needed for brute force processing power)
At what point does overlap simply not achieved any further gain. I can overlap to any extreme degree. Thus is 99% overlap does the program simply eliminate redundancy or does it place an inordinate burden on the processing calculation without much model gain above some percentage overlap ?