Abstract
In this paper we will introduce a new class of median type filters to be used in image processing. Two specific applications will be discussed: filters for edge detection and filters for noise reduction without blurring the edges and without losing fine details of the image. The filters are nonlinear, due to the median operation, in a very desirable way: the filters automatically adapt to the local image properties. When an edge is present, the filters have an excellent edge response. On flat areas the edge response is traded to good noise attenuation, thus, minimizing the appearance of spurious edges. In edge detection the new filters are shown to have better ability to detect real edges and lower probability to give false edges on flat regions of noisy images than conventional linear edge detectors.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
