I am using single pictures/frames to scan a complex object using my Miraco. I took perhaps 50 pictures and then Fused and Meshed, at which point I realized that I had missed some places. I went back and added another 15 pictures, but when I tried to Fuse and Mesh, the software created a new model, rather than adding to the first model.
How do I add frames so that I can take a somewhat good scan and complete it to make it really good?
By the way, I own 2 or 3 RevoPoint scanners plus a Creality scanner, and have used photogrammetry for years. The Miraco was able to scan a complex object more quickly and more reliably than any other scanner technology I own. The structured light scans were a game changer for me.
I tried the Resume Scan feature, but no matter how many additional frames I added aimed at the places where the mesh was incomplete, the mesh did not change or improve. Suggestions?
Also, it is confusing to see both “<- Scan” and “Resume Scan” buttons. If “<- Scan” is going to create a new scan, then it should say “<- New Scan”. The difference between the two options is not obvious.
Imagine I am scanning a cylinder sitting on a board, with the circular cross section on the board. Imagine that most of the board and cylinder have been scanned successfully, but in one spot, the place where the board and cylinder meet has no scan data. I restart the scan and take 5 frames of the place where the scan data is missing. I then repeat the Fuse and Mesh, and the model has not changed. The 5 new frames have not filled in the missing data.
I cannot add new frames in a way that fills in the missing data.
Regarding support email , there is also second option to merge the 2 pointclouds together if difficult to scan at once , but if possible I always prefer to capture as much I can in one run making sure not overlapping too much the scan .
It always depends of the object type .
When you scanning one object from many angles , please use pause button while changing the angle of the object then start again scan 360 degrees angle and then pause change angle and scan again .
It is better than resuming scan , remember resuming scan need to have at least 20% visible preview angle to be able to pick up the tracking .
Please check the Tutorial section if the forum , there is lots of videos that can help you out .
Can you explain “remember resuming scan need to have at least 20% visible preview angle to be able to pick up the tracking”? This is hard to “remember” given that I can find NO written documentation about this feature.
Are you saying that I would need to have enough frames in the Resume part in order to allow those frames to be related to the original scan/frames? This is puzzling: why not just dump all of the frames together?
Hi Jon , structured light scanner relay on object’s features for tracking , it is like a face recognition, it need to recognize the object to be able to align the frames for that reason it needs at least 20% of the scanned object to align it proper .
This is how this scanning technology works . It don’t take pictures it actually scanning the surface and each object feature is a tracking points.
Frames are like a shingles aligned one by one , so to continue scanning it need to recognize the same feature already scanned vs the feature on the object to align proper the new frames to the old frames .