You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@jian-dong and team,
We tried to implement the Orbbec Femto Mega camera system in the swine barn for data collection. We tested the system in the laboratory and it was working fine. When we implemented our system in the swine barn, we observed a noisy depth image. We have the following setup and environmental conditions in there.
Installed camera at the ceiling pointing downward with an oblique angle.
We have used a fresh camera with firmware version 1.2.9, so we believe there is no issue with firmware and dirty lens.
There is good environment, which we can verify from the RGB image captured by the same Femto Mega.
Our swine research center is also pretty clean than intensive commercial farm.
Well enough power supply to the camera through its own Adapter.
Camera is connected using both USB and Ethernet but issue is persisted on both.
Checked the IR camera too, there are some noises. It seems there are some tiny flies or dust but I trust this is hardly visible from the bare eye.
Therefore, we couldn't collect a good data there. Can you please suggest anything to improve the depth data quality? Or is it hardware or principle (ToF) constraints? Here, I have attached some sample images. The depth values in those black blobs are 0 and they are pretty random and severe to destroy the depth images.
The text was updated successfully, but these errors were encountered:
@jian-dong and team,
We tried to implement the Orbbec Femto Mega camera system in the swine barn for data collection. We tested the system in the laboratory and it was working fine. When we implemented our system in the swine barn, we observed a noisy depth image. We have the following setup and environmental conditions in there.
Therefore, we couldn't collect a good data there. Can you please suggest anything to improve the depth data quality? Or is it hardware or principle (ToF) constraints? Here, I have attached some sample images. The depth values in those black blobs are 0 and they are pretty random and severe to destroy the depth images.
The text was updated successfully, but these errors were encountered: