isaac_ros_image_proc

Source code on GitHub.

Quickstart

Set Up Development Environment

1. Set up your development environment by following the instructions in getting started.

2. Clone isaac_ros_common under ${ISAAC_ROS_WS}/src.

   cd ${ISAAC_ROS_WS}/src && \
      git clone https://github.com/NVIDIA-ISAAC-ROS/isaac_ros_common.git
   

3. (Optional) Install dependencies for any sensors you want to use by following the sensor-specific guides.

   Warning

   We strongly recommend installing all sensor dependencies before starting any quickstarts. Some sensor dependencies require restarting the Isaac ROS Dev container during installation, which will interrupt the quickstart process.

Datasets

Download Quickstart Assets

1. Download quickstart data from NGC:

   Make sure required libraries are installed.

   sudo apt-get install -y curl tar
   

   Then, run these commands to download the asset from NGC.

   NGC_ORG="nvidia"
   NGC_TEAM="isaac"
   NGC_RESOURCE="isaac_ros_assets"
   NGC_VERSION="isaac_ros_image_proc"
   NGC_FILENAME="quickstart.tar.gz"
   
   REQ_URL="https://api.ngc.nvidia.com/v2/resources/$NGC_ORG/$NGC_TEAM/$NGC_RESOURCE/versions/$NGC_VERSION/files/$NGC_FILENAME"
   
   mkdir -p ${ISAAC_ROS_WS}/isaac_ros_assets/${NGC_VERSION} && \
       curl -LO --request GET "${REQ_URL}" && \
       tar -xf ${NGC_FILENAME} -C ${ISAAC_ROS_WS}/isaac_ros_assets/${NGC_VERSION} && \
       rm ${NGC_FILENAME}
   

Build isaac_ros_image_proc

Binary PackageBuild from Source

1. Launch the Docker container using the run_dev.sh script:

   cd ${ISAAC_ROS_WS}/src/isaac_ros_common && \
   ./scripts/run_dev.sh
   

2. Install the prebuilt Debian package:

   sudo apt-get install -y ros-humble-isaac-ros-image-proc
   

Run Launch File

RosbagRealSense CameraHawk Camera

1. Continuing inside the Docker container, install the following dependencies:

   sudo apt-get install -y ros-humble-isaac-ros-examples
   

2. Run the following launch file to spin up a demo of this package using the quickstart rosbag:

   Resize:

   ros2 launch isaac_ros_examples isaac_ros_examples.launch.py launch_fragments:=resize
   

   Color Conversion:

   ros2 launch isaac_ros_examples isaac_ros_examples.launch.py launch_fragments:=color_conversion interface_specs_file:=${ISAAC_ROS_WS}/isaac_ros_assets/isaac_ros_image_proc/quickstart_interface_specs.json
   

   Crop:

   ros2 launch isaac_ros_examples isaac_ros_examples.launch.py launch_fragments:=crop interface_specs_file:=${ISAAC_ROS_WS}/isaac_ros_assets/isaac_ros_image_proc/quickstart_interface_specs.json
   

   Rectify:

   ros2 launch isaac_ros_examples isaac_ros_examples.launch.py launch_fragments:=rectify_mono interface_specs_file:=${ISAAC_ROS_WS}/isaac_ros_assets/isaac_ros_image_proc/quickstart_interface_specs.json
   

   Flip:

   ros2 launch isaac_ros_examples isaac_ros_examples.launch.py launch_fragments:=flip
   

2. Open a second terminal inside the Docker container:

   cd ${ISAAC_ROS_WS}/src/isaac_ros_common && \
   ./scripts/run_dev.sh
   

3. Run the rosbag file to simulate an image stream:

   ros2 bag play --loop ${ISAAC_ROS_WS}/isaac_ros_assets/isaac_ros_image_proc/quickstart --remap /hawk_0_left_rgb_image:=/image_raw /hawk_0_left_rgb_camera_info:=/camera_info
   

Note

For RealSense camera package issues, please refer to the section here.

Other supported cameras can be found here.

For camera calibration, please refer to this guide.

Visualize Results

1. Open a new terminal inside the Docker container:

   cd ${ISAAC_ROS_WS}/src/isaac_ros_common && \
      ./scripts/run_dev.sh
   

2. Observe the image_proc output on the terminal with the command:

   Resize:

   ros2 run image_view image_view --ros-args --remap image:=resize/image
   

   Color Conversion:

   ros2 run image_view image_view --ros-args --remap image:=image_mono
   

   Crop:

   ros2 run image_view image_view --ros-args --remap image:=crop/image
   

   Rectify:

   ros2 run image_view image_view --ros-args --remap image:=image_rect
   

   Flip:

   ros2 run image_view image_view --ros-args --remap image:=image_flipped
   

Overview

The isaac_ros_image_proc package offers functionality for rectifying/undistorting images from a monocular camera setup, resizing the image, and changing the image format. It largely replaces the image_proc package, though the image format conversion facility also functions as a way to replace the CPU-based image format conversion in cv_bridge. The rectify node can also resize the image; if resizing is not needed, specify the output width/height same as input.

Available Components

CropNode

ROS Parameters

ROS Parameter	Type	Default	Description
input_width	int64_t	0	The width of the input image. This is a mandatory parameter and needs to be set to a non-zero positive number.
input_height	int64_t	0	The height of the input image. This is a mandatory parameter and needs to be set to a non-zero positive number.
crop_width	int64_t	0	The width of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.
crop_height	int64_t	0	The height of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.
crop_mode	string	""	Mode of operation to perform cropping. Valid values are - CENTER, LEFT, RIGHT, TOP, BOTTOM, TOPLEFT, TOPRIGHT, BOTTOMLEFT, BOTTOMRIGHT, and BBOX. Except BBOX, region of interest calculation is performed based on input and cropped image dimensions.
roi_top_left_x	int64_t	0	X value of the top left corner of a region of interest. This is only used when crop_mode is set to BBOX.
roi_top_left_y	int64_t	0	Y value of the top left corner of a region of interest. This is only used when crop_mode is set to BBOX.
num_blocks	int	40	The number of pre-allocated memory output blocks, should not be less than 40.

ROS Topics Subscribed

ROS Topic	Interface	Description
image	NitrosImage	The input image that needs to be cropped.
camera_info	NitrosCameraInfo	The corresponding camera_info of the input image.

ROS Topics Published

ROS Topic	Interface	Description
crop/image	NitrosImage	Cropped image.
crop/camera_info	NitrosCameraInfo	The corresponding camera_info of the cropped image.

ImageFlipNode

ROS Parameters

ROS Parameter	Type	Default	Description
flip_mode	string	BOTH	Axis to be use for flip operation. Valid values are - HORIZONTAL, VERTICAL, and BOTH.

ROS Topics Subscribed

ROS Topic	Interface	Description
image	NitrosImage	The input image that needs to be flipped.

ROS Topics Published

ROS Topic	Interface	Description
image_flipped	NitrosImage	Flipped image.

ImageFormatConverterNode

ROS Parameters

ROS Parameter	Type	Default	Description
image_width	int64_t	1280	The width of the input image. This is a mandatory parameter and needs to be set to a non-zero positive number.
image_height	int64_t	720	The height of the input image. This is a mandatory parameter and needs to be set to a non-zero positive number.
encoding_desired	string	""	Target color encoding to convert to. Valid values are - img_encodings::RGB8, img_encodings::RGB16, img_encodings::BGR8, img_encodings::BGR16, img_encodings::MONO8, img_encodings::MONO16, img_encodings::NV24, and nv12.

ROS Topics Subscribed

ROS Topic	Interface	Description
image_raw	NitrosImage	The input image data.

ROS Topics Published

ROS Topic	Interface	Description
image	NitrosImage	The output image data.

ImageNormalizeNode

ROS Parameters

ROS Parameter	Type	Default	Description
mean	std::vector<double>	{0.5, 0.5, 0.5}	The mean of the image pixels per channel that will be used for normalization.
stddev	std::vector<double>	{0.5, 0.5, 0.5}	The standard deviation of the image pixels per channel that will be used for normalization.

ROS Topics Subscribed

ROS Topic	Interface	Description
image	NitrosImage	The input image data.

ROS Topics Published

ROS Topic	Interface	Description
normalized_image	NitrosImage	The output image data.

PadNode

ROS Parameters

ROS Parameter	Type	Default	Description
output_image_width	uint16_t	1200	The width of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.
output_image_height	uint16_t	1024	The height of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.
padding_type	string	CENTER	Describes the padding mode. Valid values are - CENTER, TOP_LEFT, TOP_RIGHT, BOTTOM_LEFT, and BOTTOM_RIGHT.
border_type	string	CONSTANT	Describes how border values are calculated. Valid values are - CONSTANT, REPLICATE, REFLECT, WRAP, and REFLECT101. More info can be found here.
border_pixel_color_value	std::vector<double>	{0.0, 0.0, 0.0, 0.0}	RGBA color value for border color.

ROS Topics Subscribed

ROS Topic	Interface	Description
image	NitrosImage	The input image data.

ROS Topics Published

ROS Topic	Interface	Description
padded_image	NitrosImage	The output image data.

RectifyNode

ROS Parameters

ROS Parameter	Type	Default	Description
output_width	uint16_t	1200	The width of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.
output_height	uint16_t	1024	The height of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.

ROS Topics Subscribed

ROS Topic	Interface	Description
image	NitrosImage	The input image that needs to be rectified.
camera_info	NitrosCameraInfo	The corresponding camera_info of the input image.

ROS Topics Published

ROS Topic	Interface	Description
crop/image	NitrosImage	Rectified image.
crop/camera_info	NitrosCameraInfo	The corresponding camera_info of the rectified image.

ResizeNode

ROS Parameters

ROS Parameter	Type	Default	Description
input_width	int64_t	0	The width of the input image. This is a mandatory parameter and needs to be set to a non-zero positive number.
input_height	int64_t	0	The height of the input image. This is a mandatory parameter and needs to be set to a non-zero positive number.
output_width	int64_t	1080	The width of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.
output_height	int64_t	720	The height of the output image. This is a mandatory parameter and needs to be set to a non-zero positive number.
keep_aspect_ratio	bool	false	When enabled, aspect ratio of the original image will be preserved by adding black border.
encoding_desired	string	""	This value is used to mark the preference of output color format.
disable_padding	bool	false	When set to true, keep_aspect_ratio is ignored and no padding is added to the output image.
num_blocks	int	40	The number of pre-allocated memory output blocks, should not be less than 40.

ROS Topics Subscribed

ROS Topic	Interface	Description
image	NitrosImage	The input image that needs to be resized.
camera_info	NitrosCameraInfo	The corresponding camera_info of the input image.

ROS Topics Published

ROS Topic	Interface	Description
resize/image	NitrosImage	Resized image.
resize/camera_info	NitrosCameraInfo	The corresponding camera_info of the resized image.

Limitation: isaac_ros_image_proc nodes require even number dimensions for images.