Saliency based automatic image cropping using support vector machine classifier

Image cropping is the process for removing the unnecessary contents from an image to improve visual composition. In this paper, we present a learning based approach for automatic cropping using saliency map. Support vector machine (SVM) model is employed to determine the cropping window. Distinct image features extracted from training set are utilized to train SVM. Proposed method enhances classical saliency based cropping technique using modified approach. We have validated our algorithm on training as well as testing dataset. Experimental results show the effectiveness of proposed method that can be useful in many applications.