Looking at the markers it looks like the resolution was set on a high point of the distance was out of the good zone
You should get much better results than this .
Hi Robert, after sending your issue to our R&D team, they think that your scanning scene setup is done very well!
Here are the reasons they think might be causing the uneven surfaces:
You only took 32 photos, which is too few and may affect the accuracy of the scanned model. You can try taking more photos before processing the calculation.
You scanned 4 parts and then merged them. The results of merging multiple parts may not align well, causing point cloud overlap and affecting surface smoothness.
If the exposure of the depth camera is not properly controlled during scanning, being too far from the car with the MIRACO Plus can result in underexposure or being too close can lead to overexposure, both of which can cause this uneven phenomenon.
Here are some scanning suggestions they have provided for you:
You can try taking more photos, around 100 or more, and directly generate a larger global marker frame. This way, you can try scanning the entire car at once or merging only 2 or 3 parts.
When scanning the entire car based on the global marker frame, it is best to keep the scanner stable and move slowly to avoid poor point cloud alignment due to instability. You can even use single-shot mode to scan the entire car region by region to avoid shaking.
If the scanning result of a certain part is not good, you can rescan that part using the same global marker frame. Then, directly crop and merge it with the previously scanned model (at this point, the exported point cloud model position and orientation will be aligned directly, and no additional alignment is needed), replacing the problematic areas. This should also be feasible.
Regarding the scanning distance, position the scanner at the recommended distance between 0.8-1.3 meters until the coded targets are clearly visible. Try to keep the depth camera displaying gray or with some small areas of blue.
When I look at your scan I see that the point distance is set too low or you scanned with greater distance that it should , the details have low accuracy so or you scanned at greater distance than GOOD or your fusing/ meshing setting was on low .
It is hard to pin point it the issue just from looking at the image.
Structured light scanners need to be at Excellent/Good distance to provide the best sharpness and quality of the capture and should be consistent on each section you scanning to preserve the same accuracy.
Scanning a car can take sometimes to 12 hours to capture the all sections , don’t try to capture too big section ensuring you keep the same distance while capturing it.
Hi! The lower „2" only impacts the color capturing exposure, but would be grayed out unless you activated color capture.
Exposure “1” is “the” important exposure. Next to keeping the optimal distance this is crucial for good scans. Red means overexposed. Blue means underexposed. Ideally, you want to have just a little bit of red and very little blue. I always try “auto” first and look how the auto adjustment reacts during the scan. It normally adjusts just fine for surfaces of uniform properties (which should be the case for your car body). If it doesn’t react to my expectation, I set the exposure manually according to the rule above (regarding red/blue).
You suppose to use it , you can remove the bars before scanning or place them in areas where they don’t disturb the surface while scanning, the whole point of PMK of using coded targets. Other way you can just use Global Markers and don’t bother with PMK