Intel® RealSense™ Depth Camera D435 Compatability

[PPabl]

Hi,

Just wanted to know if the D435 is compatible with the Raspberry Pi 3.0?

[MartyG]

RealSense cameras cannot unfortunately be used with Raspberry Pi boards because Pi boards use USB 2.0 connectors. RealSense requires a USB 3.0 connector.

[PPabl]

Okay thanks, Is it compatible with the NVIDIA Jetson TX1?

[MartyG]

I do not have information yet on the D435's compatibility with Jetson TX1. It has been used with the older R200 camera model though.

Please paste the link below into a browser window to see the video.

youtube.com/watch?v=uTiumCtFqf8&t=2s

[PPabl]

Okay thanks, by the way is the price of the intel cameras in CAD or USD?

[MartyG]

Prices listed in the Click online store are priced in US dollars. You can do an easy conversion outside of the store at the website xe.com

[idata]

Hello DigitalWorld,

Thank you for your interest on Intel® RealSense™ Technology.The D400 series cameras are compatible with any Intel or ARM processor. The only hardware requirement is USB3.0 support.That being said, please, feel free to test your NVIDIA Jetson TX1 with the Intel® RealSense™ Depth Camera D435.I hope you find this information useful.Please do not hesitate to contact us again if you need further assistance.Best regards. Josh B.Intel Customer Support

[PPabl]

Okay thanks, and do all of the cameras come calibrated?

[MartyG]

RealSense cameras are calibrated out of the box. If you do find that you need to make adjustments, the new RealSense SDK 2.0 that the D435 camera uses comes with a tool called the Depth Quality Tool that makes it much easier than previous RealSense cameras to obtain detailed information about the camera's configuration

https://software.intel.com/en-us/realsense/sdk Intel® RealSense™ SDK 2.0 | Intel® Software

[idata]

Hello DigitalWorld,

Thank you for your reply. The camera comes ready to use and calibrated. I hope you find this information useful. Please, let us know if further assistance is required or if we can close this case.Best regards, Josh B.Intel Customer Support

[RS8]

After compiling librealsense on the TX1, I got the following error when trying to run realsense-viewer depth stream:

"Backend in rs2_open_multiple(sensor:0x7f7c000a40, profiles:0x7f7c28b090, count:2):

The requested pixel format 'Y8I ' is not natively supported by the Linux kernel and likely requires a patch!

Alternatively please upgrade to kernel 4.12 or later. Last Error: Invalid argument"

Do you have instructions on how to do this on the TX1? The latest Nvidia L4T 28.1 supports Linux kernel 4.4.15

Thanks

[MartyG]

If you are using Ubuntu, the Librealsense documentation gives the following information for updating the kernel version.

1. Update Ubuntu distribution, including getting the latest stable kernel

sudo apt-get update && sudo apt-get upgrade && sudo apt-get dist-upgrade

2. Check the kernel version using uname -r and note the exact Kernel version being installed (4.4.0-XX or 4.8.0-XX) for the next step.

3. Update OS Boot Menu and reboot to enforce the correct kernel selection with:

sudo update-grub && sudo reboot

4. When rebooting interrupt the boot process at Grub2 Boot Menu -> "Advanced Options for Ubuntu" and select the kernel version installed in the previous step.

5. Complete the boot, login and verify that the required kernel version (4.4.0-79 or 4.8.0-54 as of June 17th 2017) is in place with uname -r

[RS8]

Hi MartyG,

I was able to run the realsense-viewer on the TX1/TX2 after patching the kernel following the instructions at:

https://github.com/syedharoonalam/buildJetsonTX2Kernel https://github.com/syedharoonalam/buildJetsonTX2Kernel

I believe your instructions will only work on Ubuntu desktop not for the Nvidia TX1.

The viewer runs fine if both depth and color streams are set to 640x480, but when attempting to run both at full resolution (1280x720) the color channel starts to drop a lot of frames. Any idea how to improve this?

Thanks

[MartyG]

I'm very glad that you found a solution that works for you.

You may be interested in this message from Bjarne about reducing bandwidth.