Clustering based ACO and ABC algorithms for the shadow detection and removal

In this paper, k-means and fuzzy c-means (FCM) algorithms have been used as it is efficient in grouping based on classified data points. Then ant colony optimization (ACO) and artificial bee colony (ABC) algorithms for the detection of critical regions based on the multi-modal parameters were used. It is mainly beneficial for the detection of pixel points on the large search space. These algorithms are used with the combination of k-means and FCM algorithms. We have selected ACO algorithm as it is efficient in the rapid discovery based on feedbacks. In our case we have to recognize the image points in terms of correlated pixels based on the feedbacks and clustering thresholds of the related pixels. So, ACO may be beneficial in this case and also avoids premature convergence. The ABC algorithm has been considered for the fast convergence and efficient in outlier masking. The k-means and fuzzy c-means-ant colony optimization (KFCM-ACO) and k-means and fuzzy c-means-artificial bee colony (KFCM-ABC) algorithms are used to optimize the search process and to find the segmented and associated pixels for shadow detection. The comparison of the approaches clearly depicted that the approaches have less error rates and higher accuracy.