The electric-field, caused by corrosion and prevention of corrosion method was used to overcome the deficiency for acoustic detection system in underwater target detection. The research on electric-field detection was based on electric-field sensor linear array and digital signal processing technology. Firstly, the electric-field distribution feature was simulated by electric-dipole model in shallow sea. The electric-field sensor vertical line arrays were laid regularly to obtain electric-field signal. Then, the generalized likelihood ratio (GLRT) method was used to process target detection and parameter estimation. Finally, the simulation experiment was done and indicated that the method could accurately realize parameter estimation on target initial position, velocity and direction, etc all when the signal-to-noise ratio (SNR) was greater than 0dB. The error became larger as the SNR reduced. The GLRT still could detect target when the SNR reduced to-10dB and partial parameter estimation error became larger.
An Unmixing-Based Network for Underwater Target Detection from Hyperspectral Imagery
