Abstract
In current remote sensing literature, the problems of sea-land segmentation and ship detection (including in-dock ships) are investigated separately despite the high correlation between them. This inhibits joint optimization and makes the implementation of the methods highly complicated. In this paper, we propose a novel fully convolutional network to accomplish the two tasks simultaneously, in a semantic labeling fashion, i.e., to label every pixel of the image into 3 classes, sea, land and ships. A multi-scale structure for the network is proposed to address the huge scale gap between different classes of targets, i.e., sea/land and ships. Conventional multi-scale structure utilizes shortcuts to connect low level, fine scale feature maps to high level ones to increase the network’s ability to produce finer results. In contrast, our proposed multi-scale structure focuses on increasing the receptive field of the network while maintaining the ability towards fine scale details. The multi-scale convolution network accommodates the huge scale difference between sea-land and ships and provides comprehensive features, and is able to accomplish the tasks in an end-to-end manner that is easy for implementation and feasible for joint optimization. In the network, the input forks into fine-scale and coarse-scale paths, which share the same convolution layers to minimize network parameter increase, and then are joined together to produce the final result. The experiments show that the network tackles the semantic labeling problem with improved performance.
Highlights
Remote sensing imagery is one important solution to maritime surveillance, because of its wide field of view, satisfying spatial resolution and update frequency
To ameliorate the trade-off between the network’s receptive field and the graphics processing unit (GPU) memory requirement, we introduce a novel multi-scale structure for the semantic labeling network, which greatly increases the receptive field of the network, with only a small number of parameter increase
In the remote sensing imagery the land area features most complex objects, we find the classification of these objects will not contribute to the performance of the task
Summary
Remote sensing imagery is one important solution to maritime surveillance, because of its wide field of view, satisfying spatial resolution and update frequency. Two of the most important tasks in understanding remote sensing images that is maritime-related, would be sea-land segmentation and ship detection. We propose to address the sea-land segmentation and ship detection at the same time, with a deep neural network, in a semantic labeling perspective. Since recent research shows that neural networks based on everyday object knowledge can have satisfactory performance on remote sensing imagery [41], the application of deep learning in remote sensing imagery is promising In both sea-land segmentation and ship detection tasks, semantic labeling using deep networks shows great potential.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have