An Efficient Approach for the Removal of Bipolar Impulse Noise using Median Filter

Background/Objectives: A new decision based algorithm for the removal of salt and pepper noise, also known as bipolar impulse noise. Methods/Statistical Analysis: Comparison of matrices such as 3x3, 5x5, 7x7, 9x9 is performed using median filtering techniques. Results/Findings: The result obtained, it is found that 5x5 is better to produce the clear noise free image with high degree of salt and pepper noise elimination compared to previously proposed method. Conclusion/ Application: It can be concluded that by comparing 3x3, 5x5, 7x7, 9x9 corrupted matrixes of an image using standard median filtering techniques proves that 5x5 is better than all other matrix combinations. This helps to eliminate the salt and pepper noise of corrupted pixel and provides better noise elimination capability. The amount of noise eliminated can be estimated by considering the PSNR values.