Gaussian Noise:
Because of its mathematical tractability in both the spatial and frequency domains, Gaussian (also called normal) noise models are used frequently in practice.
In fact, this tractability is so convenient that it often result. Gaussian models being used in situation in which they are marginally applicable at best.
The PDF of Gaussian random variable, z, is given by
Where,
z = Gray levels
µ = Mean value of z
σ = Standard deviation
= Variance of z
Fig. gaussian Noise
A plot of this function is shown in fig. When z is described by equation(1), approximately 70% of its values will be in the range [(µ-σ),(µ+σ)] and about 95% will be in the rande [(µ-2σ),(µ+2σ)].
Gaussion Noise is generated due to
· Electronic circuit noise
· In sensor noise due to poor illumination
· Sensor noise due to high temperature.