Chapter 6 - Quantum-inspired multi-objective simulated annealing for bilevel image thresholding

Abstract

A new quantum-based multi-objective simulated annealing technique for bilevel thresholding is developed. The proposed approach is applied on grayscale images and two steps are followed in succession to find the optimal threshold value of the input image. Firstly, two different objective functions (generally of conflicting type) are used to determine the set of nondominated solutions; thereafter, an objective function, called “Huang’s method,” is applied on these nondominated solutions to find the optimal threshold value for bilevel image thresholding. The proposed method is designed by the mingling of the basic quantum properties and a popular metaheuristic algorithm, called “simulated annealing,” on the core of multi-objective optimization. The same set of experiments was conducted on the basic multi-objective simulated annealing method as the part of a comparative analysis. The best fitness value over 40 runs, the optimal threshold value, and the computational time (in seconds) corresponding to the best result are reported for each method. The mean and the standard deviation of the objective values are also reported to judge the accuracy and robustness of each method. Finally, the preeminence of the proposed method over its counterpart is established in terms of the peak signal to noise ratio. The experimental results demonstrate that the proposed approach is superior to the equivalent classical method in all respects.