Jetson TX1 Installation
Note: The NVIDIA Jetpack Version 2.2 is currently unsupported. Jetpack version 2.1 needs to be used when configuring your TX1 for OpenPTrack.
OpenPTrack is currently being ported and tested to work on the Jetson TX1. As the TX1 has an ARM processor, the installation process and libraries deviate from the Master branch installation process.
N.B.: Currently, the onboard USB 3.0 port on the TX1 cannot support the required bandwidth for the Kinect v2. A workaround for this is to use an external PCIe USB 3.0 card with a Renesas uPD720202 chipset, such as this [Rosewill] (http://www.newegg.com/Product/Product.aspx?Item=N82E16815166038&nm_mc=KNC-GoogleAdwords-PC&cm_mmc=KNC-GoogleAdwords-PC-_-pla-_-Add-On+Cards-_-N82E16815166038&gclid=CjwKEAiA9c-2BRC_vaaJ0Ybps30SJABlqxDeiK3I1DWoFdOX1ScVoedT6LmASITELVDelDJRDR31WRoCvU_w_wcB&gclsrc=aw.ds).
N.B.2: Intrinsic calibration cannot be completed on the TX1. This process needs to be completed on a different machine, and then the intrinsic calibration files need to be moved to the TX1.
N.B.3: The TX1 and the TK1 cannot be used to run the master processes.
N.B.4: Only the Kinect v2 can be used with the TK1; the Kinect v1 WILL NOT WORK.
**1.**Download the nVidia [Jetpack] (https://developer.nvidia.com/embedded/dlc/jetpack-l4t-2_0). A login will be required to download this file. Creating an account is free.
**2.**Now, in terminal windows, navigate to where the Jetpack file was downloaded, and run:
chmod +x JetPack*
Then:
./JetPack...run
**3.**This guide can be used to flash the TX1. When configuring the Jetpack installation for the TX1, it should look like this:
**1.**After the Jetpack installation is complete, SSH into the TX1 from another machine:
ssh -XC ubuntu@tegra-ubuntu
The default password is ubuntu, if requested.
**2.**Enabling full 3.0 performance is needed. To do this, on the TX1:
gedit /boot/extlinux/extlinux.conf
Then, the line usb_port_owner_info=0 needs to be changed to:
usb_port_owner_info=2
And the following line needs to be added as the last command in the extlinux.conf file:
usbcore.autosuspend=-1
**3.**The udev rule for the Kinect v2 needs to be added. To do this, create /etc/udev/rules.d/90-kinect2.rules by:
sudo gedit /etc/udev/rules.d/90-kinect2.rules
The following text needs to be added to the 90-kinect2.rules:
# ATTR{product}=="Kinect2"
SUBSYSTEM=="usb", ATTR{idVendor}=="045e", ATTR{idProduct}=="02c4", MODE="0666"
SUBSYSTEM=="usb", ATTR{idVendor}=="045e", ATTR{idProduct}=="02d8", MODE="0666"
SUBSYSTEM=="usb", ATTR{idVendor}=="045e", ATTR{idProduct}=="02d9", MODE="0666"
**4.**Now, create the MaxPerformance scripts, and give them execution permissions:
**4.1.**Create /usr/bin/CPUSetterTX1 with the following text:
sudo gedit /usr/bin/CPUSetterTX1
With the text:
#!/bin/bash
# online all CPUs - ignore errors for already-online units
echo "onlining CPUs: ignore errors..."
for i in 0 1 2 3 ; do
echo 1 > /sys/devices/system/cpu/cpu${i}/online
done
echo "Online CPUs: `cat /sys/devices/system/cpu/online`"
# set CPUs to max freq (perf governor not enabled on L4T yet)
echo userspace > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
cpumax=`cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_frequencies | awk '{print $NF}'`
echo "${cpumax}" > /sys/devices/system/cpu/cpu0/cpufreq/scaling_setspeed
for i in 0 1 2 3 ; do
echo "CPU${i}: `cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq`"
done
Now, save CPUSetterTX1 and give it execution permissions:
sudo chmod +x /usr/bin/CPUSetterTX1
**4.2.**Create /usr/bin/GPUSetterTX1 with the following text:
sudo gedit /usr/bin/GPUSetterTX1
With the text:
#!/bin/bash
# max GPU clock (should read from debugfs)
cat /sys/kernel/debug/clock/gbus/max > /sys/kernel/debug/clock/override.gbus/rate
echo 1 > /sys/kernel/debug/clock/override.gbus/state
echo "GPU: `cat /sys/kernel/debug/clock/gbus/rate`"
# max EMC clock (should read from debugfs)
cat /sys/kernel/debug/clock/emc/max > /sys/kernel/debug/clock/override.emc/rate
echo 1 > /sys/kernel/debug/clock/override.emc/state
echo "EMC: `cat /sys/kernel/debug/clock/emc/rate`"
Now, save GPUSetterTX1, and then give it execution permissions:
sudo chmod +x /usr/bin/GPUSetterTX1
**4.3.**Create /usr/bin/MaxPerformanceTX1 with the following text:
sudo gedit /usr/bin/MaxPerformanceTX1
With the text:
#!/bin/bash
NoAutosuspending
# turn on fan for safety
echo "Enabling fan for safety..."
if [ ! -w /sys/kernel/debug/tegra_fan/target_pwm ] ; then
echo "Cannot set fan -- exiting..."
fi
echo 255 > /sys/kernel/debug/tegra_fan/target_pwm
CPUSetterTX1
GPUSetterTX1
Now, save MaxPerformanceTX1, and then give it execution permissions:
sudo chmod +x /usr/bin/MaxPerformanceTX1
**5.**If installed, uninstall the opencv4tegra packet:
sudo apt-get remove --purge libopencv4tegra*
**6.**Install utilities and permits for Ubuntu Universe and Multiverse:
sudo apt-add-repository universe
sudo apt-add-repository multiverse
sudo apt-get update
sudo apt-get install -y bash-completion command-not-found locate git gitg vim
**7.**Install libopencv-dev packet:
sudo apt-get install libopencv-dev
**8.**Install and configure ROS base system and Point Cloud Library:
sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu $(lsb_release -sc) main" > /etc/apt/sources.list.d/ros-latest.list'
sudo apt-key adv --keyserver hkp://pool.sks-keyservers.net --recv-key 0xB01FA116
sudo apt-get update
sudo apt-get install -y ros-indigo-desktop
sudo apt-get install -y python-rosinstall
sudo add-apt-repository -y ppa:v-launchpad-jochen-sprickerhof-de/pcl
sudo apt-get update
sudo apt-get install -y libpcl-1.7-all
sudo apt-get install -y ros-indigo-compressed-depth-image-transport ros-indigo-compressed-image-transport ros-indigo-pcl-ros ros-indigo-camera-info-manager ros-indigo-driver-base ros-indigo-calibration
sudo rosdep init
rosdep update
echo "source /opt/ros/indigo/setup.bash" >> ~/.bashrc
source ~/.bashrc
mkdir -p ~/workspace/ros/catkin/src
cd ~/workspace/ros/catkin
catkin_make --force-cmake
echo "export KINECT_DRIVER=openni" >> ~/.bashrc
echo "export LC_ALL=C" >> ~/.bashrc
echo "source /home/ubuntu/workspace/ros/catkin/devel/setup.bash" >> ~/.bashrc
**9.**Install and configure NTP per the installation guide.
**10.**Install Libfreenect2 and Max Performance Scripts.
sudo reboot
After rebooting, SSH back into the TX1:
ssh -X [email protected]
Install the required scripts:
cd
sudo apt-get install -y build-essential libturbojpeg libtool autoconf libudev-dev cmake mesa-common-dev freeglut3-dev libxrandr-dev doxygen libxi-dev libjpeg-turbo8-dev
git clone https://github.com/openptrack/libfreenect2.git
cd libfreenect2
git checkout jetson-dev
cd depends
./install_deps.sh
sudo ln -s /usr/lib/arm-linux-gnueabihf/libturbojpeg.so.0.0.0 /usr/lib/arm-linux-gnueabihf/libturbojpeg.so
cd ..
cd jetson_scripts
chmod +x *
sudo cp * /usr/bin
**11.**Change line 118 of examples/protonect/CMakeLists.txt from:
FIND_LIBRARY(TEGRA_JPEG_LIBRARY jpeg PATHS /usr/lib/arm-linux-gnueabihf/tegra NO_DEFAULT_PATH)
to:
FIND_LIBRARY(TEGRA_JPEG_LIBRARY nvjpeg PATHS /usr/lib/arm-linux-gnueabihf/tegra NO_DEFAULT_PATH)
**12.**Verify Protonect Installation.
Run (Note: this script needs to be run each time the TX1 is restarted):
sudo MaxPerformanceTX1 &> /dev/null
Any print out on the terminal is fine as long as it is not an error. Now run:
sudo Status
The terminal print out should be this:
CPUs active (it should be 0-3):
0-3
CPUs freq (it should be 2032500000):
1912500
1912500
1912500
1912500
CPU governor (it should be userspace):
userspace
userspace
userspace
userspace
GPU clock (it should be 852000000):
998400000
GPU memory clock (it should be 924000000):
1600000000
If the print out from sudo Status does not look like this, reboot your TX1 and run the command again. If the print out is still not correct, verify that all of the installation steps have been followed.
**13.**Plug in your Kinect v2 to a power source BEFORE plugging in the Kinect v2 to the external PCIe USB 3.0 port. After the Kinect v2 is powered, then proceed to connect it to the TK1 USB 3.0 port. Now run:
cd ../examples/protonect/
mkdir build && cd build
cmake ..
make -j4
../bin/Protonect
If all libraries and directions were followed properly, you will now see (with some delay, as the visualization is being delivered over SSH) three windows with the depth image, RGB image, and IR image. If you do not receive any images, and your terminal is only printing:
[DepthPacketStreamParser::handleNewData] subpacket incomplete (needed X received Y)]
then verify that your TX1 was set to max performance mode correctly and that your Kinect v2 is connected to the TX1 properly. If you are receiving USB errors, there is a possibility your Protonect library was linked to the incorrect version of libusb. To fix this, first try rebooting your TX1 and running the commands (above) again; then, connect to the correct library.
Once you are seeing the three images, kill the process and run:
sudo make install
**14.**Install IAI_KINECT2:
cd /home/ubuntu/workspace/ros/catkin/src
git clone https://github.com/OpenPTrack/iai_kinect2
cd iai_kinect2
git checkout jetson-dev
Now, Change row 86 of ~/workspace/ros/catkin/src/iai_kinect2/kinect2_bridge/CMakeLists.txt from:
/usr/lib/arm-linux-gnueabihf/tegra/libjpeg.so
to:
/usr/lib/arm-linux-gnueabihf/tegra/libnvjpeg.so
Now run:
Catkin
If you see an internal compiler error, run:
CatkinJ1
You can verify the installation by running:
sudo MaxPerformanceTX1 &> /dev/null
roslaunch kinect2_bridge kinect2_bridge_ir.launch
The Terminal print out should look similar to:
[CudaDepthPacketProcessor] avg. time: 21.5727ms -> ~46.3548Hz
[TegraJpegRgbPacketProcessor] avg. time: 17.6418ms -> ~56.6837Hz
Once this command is running, and the information verified, the process can be killed. Additionally, you can check the point cloud frequency by running:
rostopic hz /kinect2_head/depth_ir/points
(The command roslaunch kinect2_bridge kinect2_bridge_ir.launch must be running!)
**15.**Install OpenPTrack:
cd /home/ubuntu/workspace/ros/catkin/src
git clone https://github.com/OpenPTrack/open_ptrack
cd open_ptrack
git checkout jetson-dev
cd scripts
chmod +x *
./calibration_toolkit_install.sh
cd /home/ubuntu/workspace/ros/catkin/
catkin_make --pkg calibration_msgs
catkin_make --pkg opt_msgs
CatkinJ1
After OpenPTrack has been installed, follow the same procedures for configuration that are outlined in the Wiki. The next step in the process will be [Pre-Calibration Configuration] (https://github.com/OpenPTrack/open_ptrack/wiki/Pre-Calibration-File-Configuration).
N.B.: We suggest, if you are installing OpenPTrack on more than one TX1, that you clone your completed Ubuntu OpenPTrack installation, and then flash the remaining TX1s with the Operating System clone. To do this:
Cloning TX1 Ubuntu Image
**1.**Find where tegraflash.py is located (usually under the JetPack folder).
**2.**Put your device in FORCE RECOVERY mode (it must be attached via the micro-USB cable to a Ubuntu computer). It is in FORCE RECOVERY mode if lsusb shows a NVidia Corp. device.
**3.**Then, in the terminal, navigate to where your tegraflash.py is located on your Ubuntu machine (NOT the TX1) and run:
sudo ./tegraflash.py --bl cboot.bin --applet nvtboot_recovery.bin --chip 0x21 --cmd “read APP clone_image_name.img”
Restoring Clone to New TX1
**1.**Find where tegraflash.py is located (usually within the JetPack folder).
**2.**Put your new device in FORCE RECOVERY mode (it must be attached via the micro-USB cable to the computer). It is in FORCE RECOVERY mode if lsusb shows an NVidia Corp. device.
**3.**Then, in the terminal, navigate to where your tegraflash.py is located on your Ubuntu machine (NOT the TX1) and run the following:
sudo ./tegraflash.py --bl cboot.bin --applet nvtboot_recovery.bin --chip 0x21 --cmd “write APP restore_image_name.img”
Now you have a clone of your TX1 OpenpTrack installation.
N.B.: The cloning and restoring process takes slightly over an hour to complete.