Discrete wavelet transform-based ant colony optimization for edge detection

Abstract

Evolutionary Optimization has attracted many researchers to use it in solving many optimization problems that have no trivial solutions. Some of these techniques include; Genetic Algorithms (GA), Simulated Annealing (SA), Artificial Bee Colony (ABC), Ant Colony Optimization (ACO), etc.In this thesis, we first compare the performance of GA, SA, ABC and ACO algorithms in solving the well-known Travelling Salesman Problem (TSP). From the results obtained, the ACO algorithm has shown significant performance compared to the others. Hence, the performance of the ACO algorithm is tested in the 2-Dimensional (2-D) case for edge detection.In the last part of this work, the conventional 2-D ACO performance is tested in edge detection problem. It shows high performance. However, this performance can be improved further by transforming the input into different domain from the real time. Hence, we apply a Discrete-Wavelet Transform (DWT) at the input of the 2-D ACO algorithm which provides us denser and clearer images compared to the conventional ACO. Simulations show that the proposed 2-D DWT-based ACO provides very high performance compared to the conventional one, especially, when the input image is buried with noise.

Evolutionary Optimization has attracted many researchers to use it in solving many optimization problems that have no trivial solutions. Some of these techniques include; Genetic Algorithms (GA), Simulated Annealing (SA), Artificial Bee Colony (ABC), Ant Colony Optimization (ACO), etc.In this thesis, we first compare the performance of GA, SA, ABC and ACO algorithms in solving the well-known Travelling Salesman Problem (TSP). From the results obtained, the ACO algorithm has shown significant performance compared to the others. Hence, the performance of the ACO algorithm is tested in the 2-Dimensional (2-D) case for edge detection.In the last part of this work, the conventional 2-D ACO performance is tested in edge detection problem. It shows high performance. However, this performance can be improved further by transforming the input into different domain from the real time. Hence, we apply a Discrete-Wavelet Transform (DWT) at the input of the 2-D ACO algorithm which provides us denser and clearer images compared to the conventional ACO. Simulations show that the proposed 2-D DWT-based ACO provides very high performance compared to the conventional one, especially, when the input image is buried with noise.