Estimation Of Noise Level And Evaluation Of Noise Removal Results

With developing rapidly,digital image processing has infiltrated into many fields of the production,the life .science and research.But the image signal has often undergone the disturbance of all kinds of the noise in the process of producing.transfering,and noting. If the image has the large noises,it can influence the visual effect and the effective information detection.So a good image processing system should take the lessening of the former noise as the important step,noise removal has been the keystep of the image processing.Nowadays,our methonds of noise removal estimate or control the results of noise removal with the help of the image noise level in apply,but the noise level is unknown and it needs us to know with the method of the Probability and Statistics .In our common use of the noise level estimating method, the influence of pixels on the edge has not been considered,which mades the estimated value inexactly.In the dissertation,the author will estimate and compute again by introducing the edge detection.By a great deal of experimental datus,we can see that the veracity of the noise level has been increasing more.Basing on this,the author attempt to bring forward a new noise removal evaluating method,it has changed the case which we need recur to the standard image that can not be known in the evaluating method,and it explains that the evaluating method has more practicability and applicability.