Condensed-Sphere Ship Detection on Space Borne Optical Image Using Machine Learning Approach

Ship detection in space borne remote sensing images is of fundamental importance for maritime protection and other operation. This method is helpful for protection against illegal fisheries, oil discharge control, and sea pollution monitoring. Ship monitoring from satellite images provides a broad observable ground and covers large sea area and thus achieves a continuous monitoring of ship locations and movements. It is also known that optical space borne images have higher resolution and more visualized contents than other remote sensing images, which is more suitable for ship detection or recognition in the preceding applications. Machine learning technique can be trained hundred times faster than traditional neural network since its input weights and hidden node biases are randomly generated and the output weights are analytically computed. The advanced framework for ship detection is faster detection of ships than pixel domain, more reliable results to ensure accurate classification in large data volume and better utilization of information where the results are not affected by weather conditions like clouds, mist and ocean waves.

Author(s): Pavithra S, Vartika Sharma, Syed Thouheed Ahmed

Abstract | PDF

Share This Article