Anti-distortion correlation recognition of underwater beam drift laser images

Beam drift exists in underwater laser propagation. Influence of the beam drift on the images correlation recognition and the irrelevance between the random noise of inter-frame images were analyzed. The distortion-invariant matched filter design was studied, a phase-matching extraction of common features between distorted images was proposed, and the common features were used as the training samples of the synthetic discriminant function, based on which anti-distortion matching filter was designed and verified by a simulation. The Matlab simulation results show that the design of anti-distortion synthetic discriminant matching filter, combined with the pre-filtering of the image to be recognized, can effectively improve the PCE value (peak- the correlation energy ratio) of the correlation peak output and the disk values (peak identification rate).