Ultrasonic hyperspectral imaging based oil detection in sea water via spectral fringe-adjusted joint transform correlation

This paper presents a novel method of oil detection in sea water from ultrasonic hyperspectral imagery (UHI) using spectral fringe adjusted joint transform correlation (SFJTC). An UHI is generated using ultrasound attenuation coefficients of different materials. Using pure target (oil) and background (sea water) signatures and linear mixing, we have developed a SFJTC-based novel technique for oil detection in sea water using the UHI technique. We have randomly inserted targets (oil) into a sea water UHI scene with different levels of spectral variability to explore the performance of the SFJTC algorithm for different scenarios. The performance of the proposed technique is evaluated via receiver operating characteristics (ROC) curves and the area under the ROC (AUROC). Test results confirm that the proposed scheme shows good results despite the presence of a large amount of noise in the data.