Image Clustering and Classification Using Modified ABC Algorithm

This paper presents a hybrid clustering algorithm and feed-forward neural network classifier for regions of trees, shades, buildings and roads in a given image. It starts with the single step preprocessing procedure to make the image suitable for clustering. The pre-processed image is clustered using the hybrid genetic-Artificial Bee Colony (ABC) algorithm that is developed by hybridizing the ABC with FCM to obtain the effective clustering satellite image and classified using neural network. After applying Modified ABC method it is found that classification accuracy is improved for all regions.