dnn_depth_estimation

Monocular depth estimation with a DNN model, e.g. MiDaS.

Monocular depth estimation, by default with MiDaS v2.1 small, whose output is the relative inverse depth (larger is nearer, without the scale and the shift), or with a model of the same input and output.

opencv_apps	3.0.0 and later
ROS	ROS One, jazzy, kilted, lyrical and later
OpenCV	OpenCV 4.5.4 or later
Group	Image filters, the nodes of which publish the processed image in `~/image`

Usage

# ROS 2
ros2 launch opencv_apps dnn_depth_estimation.launch.xml image:=/camera/image_raw debug_view:=false
# ROS 1
roslaunch opencv_apps dnn_depth_estimation.launch image:=/camera/image_raw debug_view:=false

Try it with the demo video

Download the input video of the demo above, publish it as image with image_publisher, and run the node on it, in the terminals:

wget https://ros-perception.github.io/opencv_apps/media/clips/keybridge.mp4
# terminal 1: publish the video as image in a loop
ros2 run image_publisher image_publisher_node keybridge.mp4 --ros-args -p publish_rate:=10.0 -r image_raw:=image
# terminal 2: run the node
ros2 launch opencv_apps dnn_depth_estimation.launch.xml image:=image
# terminal 3: view the result
ros2 run rqt_image_view rqt_image_view /dnn_depth_estimation/image

In ROS 1:

# terminal 1
rosrun image_publisher image_publisher keybridge.mp4 _publish_rate:=10 __name:=camera
# terminal 2
roslaunch opencv_apps dnn_depth_estimation.launch image:=camera/image_raw

See Credits of the demo videos for the license of the video.

Subscribed topics

Topic	Type
`image`	`sensor_msgs/msg/Image`

Published topics

Topic	Type
`~/image`	`sensor_msgs/msg/Image`
`~/inverse_depth`	`sensor_msgs/msg/Image`

Parameters

See also the common parameters.

Name	Type	Default (range)	Changed at run time	Description
`backend`	string	“auto”	yes	Backend of cv::dnn: auto, cpu, opencl, opencl_fp16, cuda or cuda_fp16. auto uses CUDA when OpenCV is built with CUDA and a CUDA device is found, and the CPU otherwise.
`input_height`	int	256 (1 - 4096)	yes	Height of the network input, 256 for MiDaS v2.1 small
`input_width`	int	256 (1 - 4096)	yes	Width of the network input, 256 for MiDaS v2.1 small
`model`	string	“”	yes	Path or URL of the model (onnx) of the same input and output as MiDaS, empty for MiDaS v2.1 small. The URL is downloaded to ~/.ros/opencv_apps/models/ at the first time
`queue_size`	int	3	no, restart the node	Specify queue_size of input image subscribers

Source

src/nodelet/dnn_depth_estimation_nodelet.cpp, launch/dnn_depth_estimation.launch.xml