The Research And Application Of Sea Clutter Image Restoration Technology Based On Fuzzy Neural Network

Sea clutter usually referring to the backscattered echo coming from radar signals of ocean surface, and currently it is the main noise source of wave detection equipment and ship borne radar. When the ship borne radar detecting the target object on the ocean surface or close to over the sea, its backscattered echo received often mixed with sea clutter, which causes great interference to the detection. Therefore, the study of sea clutter suppression technology has great significance in the target detection, signal processing and system design of testing equipment.Researchers at home and abroad for sea clutter suppression technology have gradually shifted from signal processing to image processing. Since the image in the process of acquisition and transmission is inevitably interfering by noise, sea clutter suppression for the target object technology includes has two parts: one is to remove noise impact on sea clutter image; the other is to separate target object from the sea clutter image accurately. Noise removal section is the difficulty and important prerequisite of sea clutter suppression. In this paper, the image restoration technology was used, it can effectively remove noise while preserving the original features of sea clutter image and the target image.Firstly, an improved learning algorithm of fuzzy neural network was proposed. Based on the original document,by introducing a linear changed learning rate and reset weighting,an improved fuzzy neural network was proposed. The simulation examples show that the convergence speed of improved fuzzy neural network is faster than the speed of the original network. Secondly, theoretically prove the convergence of improved fuzzy neural network when the gradient of the error function has a finite number of zero. Finally, improved fuzzy neural network was used in sea clutter image restoration for the first time, and image restoration model was established. In the simulation, sea clutter images polluted by Gaussian noise and salt and pepper noise were respectively restored. Verify that the sea clutter image restoration technology based on improved fuzzy neural network is feasibility and effectiveness from the image clarity, the normalized mean square error, signal to noise ratio and peak signal to noise ratio and other aspects.