Noise Suppression Based On Non-linear Theory

Over the last three decades, the science of non-linear theory is becoming a brand-new science, and it is emphasized little by little. Different from the linear filter, the non-linear filter is an important branch of the science of non-linear theory and it use the statistical method to resolve the problem, use the discrete mathematics to denote, so it can not only remove noise effectively but also keep details of the signal sufficiently. As an effective processing technology, the non-linear filter is widely used in the digital image processing.In the chapter two, three and four of this paper, aiming at the salt-and-pepper noise , gaussian noise and hybrid noise which is make up of the both mentioned, applying some kind of non-linear theories in the aspect of noise removing, the paper brings forward some new filter algorithms at the basic of the predecessor's work.There are many filters that mentioned for the first time in the paper, such as the hybrid filter based on the median filter and the peak-and-valley filter, the improved stack filter based on omnidirectional structural elements constrains , the hybrid filter based on the morphological filter and the stack filter, an improved filter based on extremum and median value and the improved generalized morphological filter based on omnidirectional structural elements constrains and so on. The simulation experiments using the language MATLAB indicate that these new methods are effective.