Method for improving detection performance of hyperspectral target detection by using target signal exclusion

摘要

The present invention relates to a method for improving the detection performance of a method for detecting a target substance using a hyperspectral image. According to the present invention, a matched filter, which is most widely used for detecting a target using an ultrasound image, Experimental results of the adaptive cosine estimator (ACE) technique for the detection of the target signals included in the background signal by using the actual aerial ultrasonic image data and the simulated data generated by inserting the artificial target In the calculation of the covariance matrix for the background characterization, the target pixels having spectral reflectance characteristics, which are highly similar to the target spectra, are excluded from the background characterization process based on the diffraction angle, The detection performance is improved using the target signal separation which is configured so that the detection performance is significantly improved A method of detecting a target material of an image is provided.