A Ship Rotation Detection Model in Remote Sensing Images Based on Feature Fusion Pyramid Network and Deep Reinforcement Learning

Abstract

Ship detection plays an important role in automatic remote sensing image interpretation. The scale difference, large aspect ratio of ship, complex remote sensing image background and ship dense parking scene make the detection task difficult. To handle the challenging problems above, we propose a ship rotation detection model based on a Feature Fusion Pyramid Network and deep reinforcement learning (FFPN-RL) in this paper. The detection network can efficiently generate the inclined rectangular box for ship. First, we propose the Feature Fusion Pyramid Network (FFPN) that strengthens the reuse of different scales features, and FFPN can extract the low level location and high level semantic information that has an important impact on multi-scale ship detection and precise location of dense parking ships. Second, in order to get accurate ship angle information, we apply deep reinforcement learning to the inclined ship detection task for the first time. In addition, we put forward prior policy guidance and a long-term training method to train an angle prediction agent constructed through a dueling structure Q network, which is able to iteratively and accurately obtain the ship angle. In addition, we design soft rotation non-maximum suppression to reduce the missed ship detection while suppressing the redundant detection boxes. We carry out detailed experiments on the remote sensing ship image dataset, and the experiments validate that our FFPN-RL ship detection model has efficient detection performance.

Multi-Scale Remote Sensing Semantic Analysis Based on a Global Perspective

“…In an improved fully convolutional network (FCN) [30], the multi-scale contexts are obtained and fused for semantic understanding through a spatial residual inception (SRI) module, which continuously fusing multi-level features. Some others have proposed the feature fusion pyramid network (FFPN) [31] that strengthens the reuse of different scales' features; FFPN can extract the low level locations and high level semantic information that have important impacts on multi-scale ship detection and the precise location of densely parked ships. To better exploit the spatial information, for each pixel, multi-scale image patches are generated at different spatial scales, which are all centred at the same central spectrum but with smaller spatial scales.…”

“…The multi-scale concept in this study refers to the relationship between the local and the global, that is, a small local area and a large area within a certain neighborhood. However, the current multi-scale research focuses on the multi-scale feature extraction and fusion of the same training sample image [24][25][26][27][28][29][30][31]. There are few studies on how to consider the semantic analysis of variable regions in different spatial extents [32][33][34][35].…”

An Autonomous Path Planning Model for Unmanned Ships Based on Deep Reinforcement Learning

“…After enough rounds of training, the ship can find the right path or navigation strategy by itself. Fu, K et al [28] proposed a ship rotation detection model based on a feature fusion pyramid network and DRL (FFPN-RL). The ship's independent guidance and docking function is realized by applying the Dueling Q network to the inclined ship detection task.…”

MultiCAM: Multiple Class Activation Mapping for Aircraft Recognition in Remote Sensing Images

“…In recent years, with the improvement of hardware performance, deep neural networks have experienced tremendous development and have been widely applied in various fields, such as classification [7,8], detection [9,10], and segmentation [11,12]. Specifically, deep neural network have made breakthroughs in aircraft recognition in remote sensing images.…”