It me again
this is a question to everyone
I did some test with script from wiki. Now I have a problem/looking for solution, since I am doing fully automated process for one of my needs...
With script, I got 9 chunks and 9 meshes instead of one big mesh. As I understand script does the following:
1. Makes multiple chunks (my case 9) with aligned photos based on first chunk with all photos in it and all of them are aligned
2. Makes Pointcloud data inside of region box
3. Makes mesh based on pointcloud data from previous stage
4. Makes texture for mesh.
Problem: If you combine (open) meshes in for example Autodesk Maya you can see some holes (look at the picture)
I understand why we get this problem. But question is. Is there any way in Agisoft to do this a bit different to remove this problem. My thoughts were:
1. Import photos
2. Align photos
3. Make dense pointcloud
4. Make mesh from dense pointcloud
5. Make multiple chunks for example 9 (duplicate all data from original chunk: mesh) with smaller regions, like in script from wiki
6. Create texture just for part of the mesh which is inside of the bouding box. I want eacht part to have for example 4k texture
7. Important part: Export mesh but only from active region. By that I mean if you have for example mesh (from step 4) of dimension 90m*90m inside of each of 9 chunks (which are 1/9 of this area - in this case 30m*30m). In this case the active region dimensions would work like boolean and would make perfect cut. I tried this manually using Geomagic and results are epic
I divided mesh in 9 equal part usin planes.
8. happy Excalibur
So, is this kind of task possible inside of Photoscan?