Fuzzymultipass Filter For Impulse Noise Removal In Digital Images

• Main Authors: Vin Toh, Kenny Kal, Mat Isa, Nor Ashidi
• Format: Conference or Workshop Item
• Language: English
• Published: 2012
• Subjects: TK1-9971 Electrical engineering. Electronics. Nuclear engineering
• Online Access: http://eprints.usm.my/25877/1/Fuzzymultipass_Filter_For_Impulse_Noise_Removal_In_Digital_Images.pdf; http://eprints.usm.my/25877/

This paper presents a fuzzy multipass filter that effectively removes impulse noise in digital images. The proposed fuzzy inference mechanism is partly based on the well-known Fuzzy Inference Ruled by Else-action (FIRE) reasoning strategy, coupled with a newly introduced pseudo fuzzy rulebase that is represented by a set of simple logical operations. The FIRE mechanism is adopted to perform strong noise cancellation when impulses are detected, whereas the pseudo fuzzy rulebase conveniently simplifies the complicated computation and evaluation of a complex-structured rulebase in fuzzy filtering.