dalsa_genie_nano_c2420

v0.6 (David Portugal + Maria Eduarda Andrada)

ROS driver for the Dalsa Genie Nano C2420 Multispectral Camera using the GigE SDK and OpenCV. The code is based on the C++ SDK example, and it publishes two sensor_msgs/Image topics on ROS (colored and mono) and their compressed counterparts.

The driver also publishes images based on vegetation indexes NDVI, CVI and TVI, and each channel separately. All topics are published raw (sensor_msgs/Image) and compressed (sensor_msgs/CompressedImage).

Tested on Ubuntu 18.04 and ROS Melodic

Installation

1. Make sure you have the following dependencies: sudo apt install ethtool libgtk-3-dev libglade2-0 libglade2-dev

2. Register and Download the DALSA GigE-V Framework for Linux.

   • There's a copy of the v2.10.0.0157 of the framework from May 2019 here, in case you don't want to register.

3. Extract the x86 .tar to your HOME folder.

4. Run sudo ./corinstall in the ~/GigE-V-Framework_2.10.0.0157/DALSA directory.

5. Logout and login again into Ubuntu.

6. Connect the camera (in our setup we connect the camera to a switch, and make sure both are in the same ethernet network).

7. Check if you can find the camera on the network:

   • cd ~/GigE-V-Framework_2.10.0.0157/DALSA/GigeV/tools/GigeDeviceStatus

   • ./GigeDeviceStatus

   • You should now see the camera on the network:

   

8. Checking images from the camera using the SDK C++ example:

   • cd ~/GigE-V-Framework_2.10.0.0157/DALSA/GigeV/examples/genicam_cpp_demo

   • make

   • ./genicam_cpp_demo

   • When the program launches, you'll only see a black window. In the terminal, hit "G" and "Enter" and you'll now see the camera feed.

Note

You do not need to set a fixed camera IP address in the network. The driver will automatically detect it.

Compiling

cd your_work_space
catkin_make 

Example Usage

dalsa_genie_nano_c2420

Parameters

dalsa_camera_frame (string, default: dalsa_link)

The frame ID entry for the messages.

dalsa_camera_topic (string, default: dalsa_camera)

Topic to publish colored image.

camera_index (int, default: 0)

Only useful for multi-camera setups. You can specify the index of the camera that you want to run.

publish_mono (bool, default: false)

Activate publishing of the monochromatic image topic as well.

dalsa_camera_mono_topic (string, default: dalsa_camera_mono)

Topic to publish mono image (if publish_mono is true).

use_synchronous_buffer_cycling (bool, default: false)

Enable/disable buffer full/empty handling (cycling). This allows Asynchronous (false) and synchronous (true) camera stream transferring.

tune_streaming_threads (bool, default: false)

Enable/disable transfer tuning (buffering, timeouts, thread affinity). This tunes the streaming threads, such as assigning specific CPUs to threads based on affinity.

turbo_mode (bool, default: false)

Enable/disable Turbo Mode (if available) to push past the gigabit Ethernet speed ceiling, speeding up line and frame rates beyond the nominal link capacity.

Topics

dalsa_camera (sensor_msgs/Image)

RGB Image. Raw and compressed.

dalsa_camera_720p (sensor_msgs/Image)

Downscale RGB image to 720p. Raw and compressed.

dalsa_camera_mono (sensor_msgs/Image)

Mono (8UC1) Image. Raw and compressed. Only published when publish_mono is true.

dalsa_camera_mono_720p (sensor_msgs/Image)

Mono (8UC1) Image, downscaled to 720p. Raw and compressed. Only published when publish_mono is true.

ndvi (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed

Extra indexes that can be toggled on camera.yaml

cvi (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when CVI is true.

red_channel (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when Red is true.

green_channel (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when Green is true.

nir_channel (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when NIR is true.

normalized_green (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when Normalized_Green is true.

normalized_nir (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when Normalized_NIR is true.

normalized_red (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when Normalized_Red is true.

tvi (sensor_msgs/Image)

Mono (8UC1) Image. Raw and Compressed. Only published when TVI is true.

Node

roslaunch dalsa_genie_nano_c2420 dalsa_genie_nano_c2420.launch

You can change the parameters on the launch file to suit your setup.