Digital Image Processing

Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Neighborhood Based Pixel Approximation for High Level Salt and Pepper Noise Removal


Affiliations
1 Dept. of CSE, Narula Institute of Technology, Agarpara, Kolkata-109, WB, India
2 Dept. of CSE, SKFGI, Mankundu, Hooghly- 712139, WB, India
3 Netaji Subhash Engineering College (under West Bengal University of Technology), Technocity, Garia, Panchpota, Kolkata-152,WB, India
     

   Subscribe/Renew Journal


Digital gray scale images contain salt and pepper noise often due to capturing difficulties. Removal of such impulsive noise is commonly taken care by median based filters. These median based filters are inefficient when the noise level is particularly very high. This paper presents a de-noising technique where noisy pixel’s value is approximated by neighborhood pixels. This technique is expected to work in such high noise levels. Experimental results with this proposed technique show that proposed method provide better performance with respect to both Mean Square Error (MSE), Peak Signal-to-Noise Ratio (PSNR) and Image Enhancement Factor (IEF). The developed filter also performs acceptably well even at a noise level as high as 90%.


Keywords

Salt and Pepper Noise, Approximation, Image De-Noising, Filtering, Neighborhood Pixel.
User
Subscription Login to verify subscription
Notifications
Font Size

Abstract Views: 244

PDF Views: 3




  • Neighborhood Based Pixel Approximation for High Level Salt and Pepper Noise Removal

Abstract Views: 244  |  PDF Views: 3

Authors

Shubhendu Banerjee
Dept. of CSE, Narula Institute of Technology, Agarpara, Kolkata-109, WB, India
Aritra Bandyopadhyay
Dept. of CSE, SKFGI, Mankundu, Hooghly- 712139, WB, India
Rajib Bag
Dept. of CSE, SKFGI, Mankundu, Hooghly- 712139, WB, India
Atanu Das
Netaji Subhash Engineering College (under West Bengal University of Technology), Technocity, Garia, Panchpota, Kolkata-152,WB, India

Abstract


Digital gray scale images contain salt and pepper noise often due to capturing difficulties. Removal of such impulsive noise is commonly taken care by median based filters. These median based filters are inefficient when the noise level is particularly very high. This paper presents a de-noising technique where noisy pixel’s value is approximated by neighborhood pixels. This technique is expected to work in such high noise levels. Experimental results with this proposed technique show that proposed method provide better performance with respect to both Mean Square Error (MSE), Peak Signal-to-Noise Ratio (PSNR) and Image Enhancement Factor (IEF). The developed filter also performs acceptably well even at a noise level as high as 90%.


Keywords


Salt and Pepper Noise, Approximation, Image De-Noising, Filtering, Neighborhood Pixel.