This package handles the calculations for locating an underwater acoustics pinger that is pinging at a certain frequency.
To launch the main sampling and processing nodes, you can use the launch file provided. Note that this expects the Logic software to already be running and the saleae device connected via USB.
roslaunch acoustics acoustics.launch
If you are using the saleae (the hardware used to sample the hydrophones), then you will need to have X forwarding enabled (pass in -XY to the ssh command). Note that the graphics process (the saleae Logic application) will persist even after you kill the ROS node.
Refer to the section on the acoustics wrapper structure below for the specific information to send to start processing. To start an action via a GUI (without an action client) refer to this tutorial.
To use the data_sim server that will generate test data, instead of hardware sampling with the saleae, you can use the command:
roslaunch acoustics acoustics.launch sim:=true
This package contains 5 nodes to handle various acoustics tasks: the acoustics wrapper, the saleae node, the guess server, the processing server, and the data server.
The acoustics wrapper node provides an external interface (an action server) to allow client code to easily call all of the steps involved in acoustics processing without providing lower level information such as the sampling frequency or file paths for saved data.
The node itself provides an action server of type custom_msgs/AcousticsWrapper.action, which client code will use to call the action. Client code is required to provide the target frequency of the pinger (Robosub provides this depending on what course we are using), and the wrapper server will return the horizontal angle to which the pinger is located at. It will also update the client as to which step is being done via its feedback messages.
To start a GUI for this action using the builtin GUI interface, you may use the command:
rosrun actionlib_tools axclient.py /acoustics_wrapper
The saleae node interfaces with the saleae and tells it to sample data for a given duration and sample for a given amount. It saves the data to a file located in the data folder. It contains an action server using type custom_msgs/Saleae.action.
The guessing node contains code to make a guess for which octant the pinger is located in, which provides the processing node the necessary information to do its computations. The code is separated into guess_server.py and cheap_cross_cor.py to separate the ROS interface from the underlying acoustics code, to allow for the guessing code to execute without ROS. The ROS node contains an action server using type custom_msgs/AcousticsGuess.action.
The processing node contains code to actually compute the horizontal angle using cross correlation. The code is separated into processing_server.py and cross_corr_fft.py to separate the ROS interface from the underlying acoustics code, to allow for the processing code to execute without ROS. The ROS node contains an action server using type custom_msgs/AcousticsProcessing.action.
The data simulation node contains code to simulate the saleae and generate data that represents the hydrophones at a specific point. This is used for testing the guessing and processing algorithms. The code is separated into data_server.py and data_sim.py to separate the ROS interface from the underlying acoustics code, to allow for the processing code to execute without ROS. The ROS node contains an action server using type custom_msgs/AcousticsData.action.