What principle does PhotoScan use?

The principle is three-dimensional photogrammetry, and the detailed theory cannot be finished in three or two sentences. This is a set of surveying and mapping methods, which are relatively mature at present. Want to know the detailed theory can ask. I will find the information.

For a general introduction of photogrammetry, please refer to the Baidu Encyclopedia website link.

Detailed algorithm or something, you need to refer to academic papers. Specific software code, which is the core part of their software.