07-03-2011, 11:51 AM
presented by:
K.PRATHYUSHA
T.RAJ KUMAR
D.PAPARAO
G.SHIRISHA
[attachment=9668]
Bit plane
A Bit plane consists of the bits corresponding to the same significant level in all the elements
For example, the most significant bit plane is formed by considering the most significant bit(MSB) of each elements
The elements of a gray scale image have a maximum value of 255 and hence can be represented in binary domain using 8 bits
Higher order Bit planes of an image carry a significant amount of visually relevant details
Lower order Bit planes contribute more to fine details
An image can be represented by 8 Bit planes
Bit planes
Classification of denoising algorithm:
Spatial filtering Methods
1.Non linear Filters
2.linear Filters
Transform domain filtering Methods
Median filtering
Median filtering is particularly effective in the presence of impulse noise
Impulse noise is characterized by replacing a portion of an image’s pixels with random values, leaving the remainder unchanged.
In a set of ordered values, the median is the central value.
The idea is to replace the current point in the image by the median of the brightness in its neighborhood.
Unlike averaging filter, median filtering does not blur too much image details.
Advantages:
removes impulse noise
Preserves edges
Disadvantages:
Performance poor when percentage of noise pixels in the window is greater than 50% in the window
Performs poorly with Gaussian noise
3*3 window
Procedure to find median value
Initial order r=5
Select rth bit in the MSB plane
Bit selection:
Select the bits in the Bit plane whose immediate higher significant bit has a value equal to the out put of that Bit plane
If the previous Bit plane out put is 1 then the order for the next Bit plane is
r = previous order -[no of selected bits in the previous Bit plane
- no of selected bits in the present Bit plane]
otherwise
r = previous order
Select the rth bit in the present Bit plane
Repeat the procedure upto LSB
median function Flow chart
Mean filter using Bit-planes
Mean filtering is particularly effective in the presence of Gaussian noise
Gaussian noise is characterized by adding noise value to each pixel of a image
Mean filtering blurs the image
The idea is to replace the current point in the image by the mean of the brightness in its neighborhood.
Mean
Algorithm
Step 1 : Form Bit-planes for the given array of elements.
Step 2 : Count the no of 1’s in the first 3 MSB planes.
Step 3: Generate Bit-plane codes for the first 3 MSB planes based on the no of 1’s from the look up tables.
Step 4: Binary addition of the three Bit-plane codes gives the mean value for the given array.
Step 5: Replace the center pixel with the mean value.
Step 6: Repeat the algorithm to the entire image
Mean function Flow chart