I’ve shot a mapping test this morning over an averaged sized area but comprising pretty large structures/buildings. I ran the coverage steps at 88 in the capture app and also did a pass taking oblique imagery of the scene in the hopes to provide additional data for the processing.
I’m interested in extracting high resolution models for use in visual effects and so I’m trying to gain experience with your sofware in order to achieve this.
I noticed that the structure processing mode is labelled as being in “beta” but I would like to no what are the current limitations of this implementation in order for me to optimize my testing and the potential results. Is there for example an optimal area size that works best for now? Is the precision of the output model reduced on purpose for the duration of this beta. It seemed to me that I provide the software with quite a bit of imagery and was really surprised at the low resolution of the 3d geometry output.
Any pointers to improve my results? What is an average sized dataset in terms of pictures/area size that tend to produce the best results?
I uploaded 273 photos from my Phantom 3 Advanced and chose Structure. The results did not include a 3D redering and was not even a choice on the drop down menu after the processing was complete!
Thanks. But as you can see the drop down did not reveal this. This is the 1st or maybe 2nd time that I tried the Structure option. All the other times the terrain always revealed a 3D dropdown or choice.
Hi Tom, thanks for re-uploading the imagery. Looks like the 3D came through this time. Going forward, it’ll help to remove ground level imagery prior to upload.