I am trying to do the same but i get this error any solution?? [ERROR] (2021-08-05 22:18:54.972) Rtabmap.cpp:1156::process() RGB-D SLAM mode is enabled, memory is incremental but no odometry is provided. Image 3604 is ignored!
Nice. I just noticed recently how bad the Realsense cameras are though for depth. The data is extremely noisy and poor quality compared to TOF sensors. For me it is garbage. They are more affordable than Azure Kinect though. I have ordered a Realsense L515 which should hopefully be better.
@@tienhoangngoc7867 Not sure, I have tried some software filters but usually they use a lot of processing power, too slow, etc. The stereo hardware like 400 series is just not as good as time of flight hardware. TOF and Lidar measure depth directly so it is much more accurate. Even my Huawei phone with TOF sensor is more accurate but low resolution 240x180. I gave up on software methods including deep learning neural depth, etc. Better to have real measurements per pixel like with TOF sensors.
Này là mình đang chạy SLAM trên ROS, bạn có thể tham khảo trong gói ros cho dòng depth camera do Intel phát triển. github.com/IntelRealSense/realsense-ros Hiện tại mình cũng có đang làm một con robot tương tự như trong video này, github.com/IntelRealSense/realsense-ros nếu bạn quan tâm thì có thể liên lạc với mình: hoangngoctien203@gmail.com
I am trying to do the same but i get this error any solution??
[ERROR] (2021-08-05 22:18:54.972) Rtabmap.cpp:1156::process() RGB-D SLAM mode is enabled, memory is incremental but no odometry is provided. Image 3604 is ignored!
wow great ive been trying to work this out, thanks
Nice. I just noticed recently how bad the Realsense cameras are though for depth. The data is extremely noisy and poor quality compared to TOF sensors. For me it is garbage. They are more affordable than Azure Kinect though. I have ordered a Realsense L515 which should hopefully be better.
Good luck with Realsense L515 bro.
What if u combine them and use filters to reduce noisy?
@@tienhoangngoc7867 Not sure, I have tried some software filters but usually they use a lot of processing power, too slow, etc. The stereo hardware like 400 series is just not as good as time of flight hardware.
TOF and Lidar measure depth directly so it is much more accurate. Even my Huawei phone with TOF sensor is more accurate but low resolution 240x180.
I gave up on software methods including deep learning neural depth, etc. Better to have real measurements per pixel like with TOF sensors.
@@mattizzle81 What is your take on this d435+ t265 bundle? dev.intelrealsense.com/docs/depth-and-tracking-cameras-alignment
Amazimg video
Xin top
Quá hay
Hay quá, anh có thể chia sẻ code được không ạ ? Em mới bắt đầu tìm hiểu về depth cam ạ
wolfram161@gmail.com
Này là mình đang chạy SLAM trên ROS, bạn có thể tham khảo trong gói ros cho dòng depth camera do Intel phát triển.
github.com/IntelRealSense/realsense-ros
Hiện tại mình cũng có đang làm một con robot tương tự như trong video này,
github.com/IntelRealSense/realsense-ros
nếu bạn quan tâm thì có thể liên lạc với mình: hoangngoctien203@gmail.com
@@tienhoangngoc7867 chào anh, em có hỏi anh 1 số câu hỏi liên quan về topic này không ạ ?
@@quangnguyen-ji1qs được bạn, mình biết thì sẽ trả lời nhé