Segmentation of satellite image based on quantum approach and the Havrda-Charvat entropy

In this paper, we present a new segmentation founded on the theory of quantum mechanics and the Havrda-Charvat entropy. This segmentation method is a continuous improvement of the method developed in 2014 by Mechkouri et al. (J Theor Appl Inf Technol 62(2):539-545, 2014). The segmentation was performed using an unsupervised segmentation method using the hierarchical analysis of the 2D histograms, which was constructed using the wave function and Havrda-Charvat entropy. We upgrade the quality of the segmentation method by reclassifying no segmented pixels into the classes defined by the Euclidean distance and Mahalanobis distance. The proposed approach was first tested on the synthetic image; it was applied to the segmentation of the environment for the QuickBird data of a selected urban area in the region of Rabat- Sale-Kenitra, prefecture of Skhirate-Temara, Morocco. The segmentation result obtained is more satisfactory than that obtained in the previous study (Mechkouri et al. in J Theor Appl Inf Technol 62(2):539-545, 2014). The development method is very hopeful and needs to be more investigating.