A Study of Hybrid Partial Differential Equation Diffusion Image Amplification Method Using Filtering Algorithm

In this project, a hybrid partial differential equation diffusion image amplification method for double orthogonal conditions is proposed by using pulse filter, improved forward-backward diffusion filter and total variational diffusion method. The improved hybrid partial differential equation diffusion image filtering algorithm is used to strengthen the small boundary effectively. The processing ability of strong boundary is improved by using pulse filter with certain parameter limit. Using the level set method, the TV diffusion processing removes the sawtooth at the boundary and improves the smoothness of the boundary. Then, a biorthogonal filtering method based on degraded model is proposed, which can realize biorthogonal mapping in space domain, so as to ensure that the degraded extended image is invertible to degraded model. The simulation results show that this method is superior to other methods, and the obtained amplified images have good photorealism, and the boundary of weak and medium brightness is also improved. This method preserves the boundary details of the image while eliminating the noise.