Research And Application Of Stochastic Resonance In Image Denoising And Target Detection

The emergence of stochastic resonance theory confirms that the noise signal is not all harmful,it also has a useful side.From the standpoint of stochastic resonance,noise can be used to enhance the signal,which brings new ideas to image processing.Subsequently,some application methods of stochastic resonance in image denoising,edge detection and other fields have emerged.These methods have shown great potential for application of stochastic resonance in the image field,but these methods have yet to be improved.In engineering practice,it is often impossible to obtain the relevant information of the original noisy image,then,in the absence of reference image to automatically adjust the system parameters to achieve adaptive stochastic resonance image denoising is crucial.Applying the stochastic resonance theory to the target detection of low signal-to-noise ratio video can effectively detect the moving target in the video and improve the accuracy of target detection.This further expands the application of stochastic resonance.Firstly,this paper introduces the method of bistable stochastic resonance image denoising in detail.Through the simulation experiments of stochastic resonance denoising of images under different noise intensity and different noise types,the results show that bistable stochastic resonance can effectively remove image noise and improve the peak signal-to-noise ratio of the image,and the suppression of Gaussian white noise and multiplicative noise is better.Secondly,the stochastic resonance image denoising and Particle Swarm Optimization(PSO)are combined,and the no-reference quantitative indicator of noise distortion—Donoho noise standard deviation—is used as a fitness function to find the optimal parameter combination of the system.The algorithm realizes adaptive selection and optimization of parameters in stochastic resonance image processing.Experimental results show that compared with the conventional image denoising algorithm,when the amount of noise in the image is large,the adaptive stochastic resonance algorithm has better denoising effect and higher image quality.For the crystal growth image in the 8-inch silicon single crystal furnace,adaptive stochastic resonance denoising is performed to improve the image quality,thereby,the meniscus of silicon single crystal images with low signal-to-noise ratio can be accurately detected,which lays the foundation for the accurate detection of crystal diameter.Finally,aiming at the problem that the existence of noise in the low signal-to-noise ratio video causes serious interference to the detection of moving targets,a target detection algorithm based on multi-frame stochastic resonance is proposed.The experimental results show that the algorithm can effectively detect moving targets in video,and it can significantly improve the accuracy of target detection compared with hybrid Gaussian algorithm(GMM).