Ship detection in videos

• Main Author: Muhammad Mukhtar
• Other Authors: Deepu Rajan
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Computer science and engineering
• Online Access: http://hdl.handle.net/10356/72907
• Institution: Nanyang Technological University

Computer vision can be used in maritime environment to assist in ship navigation and may lead to reduction in maritime accidents. In this project, improvements were made to an existing ship detection and tracking prototype to solve the issues that the prototype had, such as false positive detections and inability to track occluded objects. Other explorations were done to improve the accuracy of the ship detections. One such exploration is water region segmentation using pixel classification into water and non-water pixels. Classification methods such as linear SVM and random forest were used and feature spaces were built using features such as pixel color features and Gabor pixel texture features.