Reserch On De-noising For The Pictures Contain Impulse Noise And Gaussian Noise

In this article, we analysised the impulse noise. According to the problem that densechunks of impulse noise in the same area is difficult to remove, we presents a de-noisingmethod can simultaneously removing impulse noise and chunks of impulse noise. In thisway the color image and the gray image are applicable. Some experiments show that thisapproach within a certain impulse noise density effect can be accepted.In this article, we analysised the Gaussian noise.For the problem on removal Gaussiannoise will lose a lot of image detail,we found a way that combination of Dual TreeComplex Wavelet de-noising and Bayesian threshold. In this way the color image and thegray image are applicable. Some experiments show that the effect of this de-noisingmethod is satisfactory and the loss of image detail is fewer in the case of Gaussian noisedensity is not high.We did some research for the problem that Gaussian noise and impulse noise occursimultaneously will make adverse effects to the method whose target is Gaussian noiseand the method whose target is impulse noise. And then find out the answer that whatwould be the impact on them. According to the answer, we presents a order that can makethe adverse effects smallest when we combine these methods. Some experiments showthat the effect of this order can be accepted.