Evaluation of a New Platform For Image Filter Evolution

This paper describes a new FPGA implementation of a system for evolutionary image filter design. Three parallel search algorithms are compared. An optimal mutation rate and the quality of three pseudo-random number generators are investigated. The efficiency of proposed system is demonstrated on the problem of removing the salt-andpepper noise with intensity of 5%, 10% and 20% and designing an edge detector which works with input images corrupted by the salt-and-pepper noise.

• Type: Conference
• Year: 2007
• Where: AHS
• Authors: Zdenek Vasícek, Lukás Sekanina
• Keywords: AHS 2007, Hardware, Optimal Mutation Rate, Parallel Search Algorithms, Pseudo-random Number Generators