A genetic approach to edge detection

A new edge detector based on Fresnel diffraction

“…Extracting these features makes contribution to higher-level visual processing like three-dimensional reconstruction, stereo motion analysis, image segmentation and image compression. Since edge detection has been an active area for more than 40 years, many effective methods are proposed such as the methods based on derivatives (Hardie and Boncelet, 1995;Chidiac and Ziou, 1999;Clavier et al, 1999;Haralick and Shapiro, 1992;Sharifi et al, 2002;El-Khamy et al, 2000;Kim et al, 2004;Bovik, 2000), optimality criteria (Canny, 1986;Sarkar and Boyer, 1991;Shen and Castan, 1992;Deriche, 1987;Demigny, 2002;Pellegrino et al, 2004), statistical procedures (Rakesh et al, 2004;Chuang and Sher, 1993;deSouza, 1983;Qie and Bhandarkar, 1996;Konishi et al, 2003b;Papachristou et al, 2000), surface fitting (Nalwa and Binford, 1986;Haralick, 1984;Sinha and Schunk, 1992;Chen and Yang, 1995), genetic algorithm (Bhandarkar et al, 1994;Caponetti et al, 1994), residual analysis-based techniques (Chen et al, 1991;Spinu et al, 1997) and support vector machine (Zheng et al, 2004).…”

Edge detection through the use of a combined genetic algorithm-linear technique approach

“…Genetic algorithms are basically optimization techniques that operate in a subset of the universe of possible solutions. A genetic algorithm approach proposed by Caponetti in [5] considers a set of edge configurations on the dimensions of the image as possible solutions for the target problem. Using the evolutionary approach of genetic algorithms, the initial set is transformed during a user-defined number of iterations when the best solution is chosen.…”