Denoising Of Remote Sensing Images Based On Partial Differential Hybrid Model With Energy Functional Theory

Images can help people observe objects more intuitively,and the application of images has become more and more widespread with the rapid development of information technology.In most image application research,high-quality digital images are often needed.However,in the process of image formation,storage and transmission,images are highly susceptible due to the influence of system and environment,among which the most common It is noise.Noise-affected images can have a significant impact on subsequent image processing.Therefore,denoising preprocessing of images is an indispensable step in image application research.Because remote sensing images can effectively and quickly acquire important information such as space features,they are more and more important in the direction of military,environmental monitoring,resource evaluation and other national key concerns,but because remote sensing images are more common than ordinary images.The other aspects are more special and diverse,which leads to the common noise pollution in real remote sensing images.Therefore,it is more important to remove the noise from remote sensing images.This paper firstly studies the effectiveness of partial differential equations in image denoising by starting from ordinary image information.This paper mainly discusses the overall variational(TV)model and the fourth-order partial differential(Y-K)model,and studies the theoretical basis of the two types of models in detail,and analyzes the two types of models through simulation experiments.Secondly,aiming at the shortcomings of the Y-K model,the improvement of the diffusion coefficient and the optimization of the laplace operator of the edge detection parameters are introduced.The improved two-order partial modification is proposed.The equation model was analyzed and simulated.Then based on the TV model and the improved model,a partial differential image denoising model based on the image energy functional extremum principle is constructed.The feasibility of the model is demonstrated by theoretical analysis,and the simulation experiment is carried out from the subjective visual effect and objective.After the image quality evaluation index proves its validity,the established hybrid model is applied to the real-time remote sensing image denoising research.The practical application results show that the partial differential hybrid model established in this paper can effectively suppress the salt and pepper and Gaussian noise contained in remote sensing images,and can effectively maintain the rich detail features in the image.