(One of Experience on Neural Networks to Detect Submarines) |
|||
| Line 1: | Line 1: | ||
| − | This is one of my experiences using Neural Networks to classify sonar signals. In the sea, there are many sonar signals from different animals, fish,adversary ships, and submarines. Our algorithms were integrated into a computer in a submarine. Sonar signals are sampled and fed into a computer to detect adversary ships or submarines. The sonar signals | + | This is one of my experiences using Neural Networks to classify sonar signals. In the sea, there are many sonar signals from different animals, fish,adversary ships, and submarines. Our algorithms were integrated into a computer in a submarine. Sonar signals are sampled and fed into a computer to detect adversary ships or submarines. The dimension of sonar signals are reduced by extracting feature vectors by using FFT, Wavelet Transform, DCT, or other features such as amplitued and phase informaiton. These feature vectors are fed into a Neural Network to train weights. After training, Neural Network classifies sonar signals in real time with robustness to the noise from sea. |
Latest revision as of 22:56, 24 April 2008
This is one of my experiences using Neural Networks to classify sonar signals. In the sea, there are many sonar signals from different animals, fish,adversary ships, and submarines. Our algorithms were integrated into a computer in a submarine. Sonar signals are sampled and fed into a computer to detect adversary ships or submarines. The dimension of sonar signals are reduced by extracting feature vectors by using FFT, Wavelet Transform, DCT, or other features such as amplitued and phase informaiton. These feature vectors are fed into a Neural Network to train weights. After training, Neural Network classifies sonar signals in real time with robustness to the noise from sea.
