When fusing, what is the area that each ZED2 camera can cover? (In other words, how should the ZED2 cameras be arranged for venues of different areas to achieve a stable Fusion effect?)
I tried some scenarios, and I found that the fusion effect was not very good.
The two ZED cameras are approximately ten meters apart, and the width of the shooting venue is about seven meters. It can be observed that the skeletons are shaking, and a certain skeleton will teleport back and forth within a certain distance.
(During the test, only one person was filmed during the calibration process, and two people were filmed after the calibration.)
The effect after fusion is as follows.
The following are the tests for other area sizes.
7x2m:
4x2m:
within 2x2m:
It can be seen that it seems that only in a small space can the skeletal fusion be relatively accurate.
This post mentions that for a venue with an area of 5x20 meters, 6 cameras are required.:
This post mentions that for a venue with an area of 15x28 meters, 10 cameras are required.:
As shown in the figure, we need to film a venue with an area of 5x5 meters. Taking into account the shooting range of the cameras and the fact that we still need to conduct some tracking of people even when they are a bit beyond the 5-meter mark, the final size should not exceed 10x7 meters. The two ZED cameras must be placed facing each other (in order to obtain the front and back images of people and use a model like YOLO for segmentation and matting).
Could you please give me some guiding suggestions to help me use the ZED 360 and Fusion smoothly?