Smart Vessel Detection using Deep Convolutional Neural Network
Iwin Thanakumar Joseph S, Raj S Benson Edwin, J Sasikala, Juliet D Sujitha · 2018
Detecting ships automatically from the satellite images is one of the major challenging constraint due to lot of disturbances because of cluttered scenes, ship sizes etc. Small object detection in satellite images like ships, boat, vessels is a challenging task. In the state of art method, numbers of features are used to enhance the accuracy of detection but that mostly happens in homogeneous environment. Few research articles deals with heterogeneous environment such as sea shore areas, harbors, islands etc. Deep learning is one of the emerging technology in the recent days especially in the field of computer vision to provide an accurate result. It paves the new channel to the entire machine learning paradigm and provides the advancement of artificial intelligence utilization and also enhances the human computer interaction in a much feasible manner. In this article the deep learning algorithms such as convolutional neural network is used with the proper hybrid and the performance is compared with the state of art method. The deep learning algorithms result excellent accuracy in terms of classification of satellite images without any manual feature extraction. Since the deep learning algorithms does not have the manual feature extraction, it provides an excellent experimental results in the classification of ships in satellite images that has complex background.