New quantum inspired meta-heuristic methods for multi-level thresholding

Thresholding is a simple, effective and popular method for image segmentation. It can be bi-level or multi-level depending on number of segments in an image. Multi-level thresholding computationally takes more time than the bi-level thresholding. To reduce the computational complexity, here we propose two quantum inspired meta-heuristic methods, namely Quantum Inspired Ant Colony Optimization and Quantum Inspired Simulated Annealing for multi-level thresholding. The basic quantum principles are coalesced with meta-heuristic approaches to design the proposed methods. The performance of the proposed methods is demonstrated in comparison with its conventional versions for two test images in terms of optimal threshold values at different levels with the fitness measure, standard deviation of the fitness measure and the computational time. It has been noticed that the Quantum Inspired meta-heuristic methods are superior in terms of computational time compare to the other methods. Finally, statistical significance test, called t-test, has performed to establish the superiority of the results.