How can raw data in sonar reflections of a target be pre-processed and provided to machine learning methods?
How can a database be created using raw sonar received from active sonar to identify target and non-target?
These raw signals contain 64 channels.
For example, some raw data is given. Also, signals related to 4 channels are plotted.
In order to get advices, can you clarify which type of sonar you use to collect data and at which stage of the sonar processing you are extracting them? Passive and active sonar processing chains are totally different!!!!! As are those of a mil-grade sonar (and subsequently a MW sonar wrt an ASW sonar...), a side-scan sonar for civil engineering, a fishery vessel sonar, an echo-sounder or an underwater location beacon... To get help and advices, please make clear the problem you want to address!
for signal classification, you can use LSTM network.
this example may help you :
https://www.mathworks.com/help/deeplearning/ug/classify-sequence-data-using-lstm-networks.html
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Computer Graphics and Animation
- 3D