In point operation, transformation function T(.) works only on single input pixel value say r In Neighbouring operation, Transformation function works on group of input values. It is a group operation. Number of input values selected for processing depends on window size.
Point operation is azero memory operation . Output pixel value at (x,y) position depends on a single input at (x,y) position. NO memory space is required to store intermediate results. In Neighborhood operation output pixel value at (x,y) position depends on many more input pixel value memory space is required to store intermediat results. In Neighborhood operation output pixel value at (x,y) position depends on many more input pixel value, therefore to calculate single output pixel value memory space is required to store the intermediate results.
Examples of Operations:
Examples of point operation include:
- Contrast Stretching Transformation
- Clipping and Thresholding
- Digital Negative
- LOG Transformation
- Power Law Transformation
- Intensity Level Slicing
- Bit Level Slicing
Examples of Neighboring Operation is Spatial filtering.
Smoothing Spatial Filters: LPF, Weighted Average filter, Trimmed Average Filter, Median, Max, Min filters.
Sharpening Filters: Derivative filters, HPF, High Boost Filters
Output Pixel Value
In point operation, the output pixel value s at (x,y) position is given by s=T(r) Where r is input pixel value at (x,y) position
In neighborhood operations , the ouput pixel value at (x,y) position is obtained by masking operation
Eg: consider 3x3 mask w and digital sub-image F,
W1 W2 W3 W4 W5 W6 W7 W8 W9
Z1 Z2 Z3 Z4 Z5 Z6 Z7 Z8 Z9
The response of filter mask at the centre position in the output image is given by,