Arbitrary-Oriented Ship Detection Based on RetinaNet for Remote Sensing Images

Abstract

Aiming to address the problems of arbitrary orientations, large aspect ratios, and dense arrangements in ship detection, an arbitrary-oriented ship detection method based on RetinaNet is proposed. Our proposed method includes a rotated RetinaNet, a refined network, a feature alignment module, and an improved loss function. First, the rotated RetinaNet achieves rotation detection by using a feature pyramid network, rotated anchors, the skew intersection-over-union (IoU), and skew non-maximum suppression (NMS). Then, the refined network and feature alignment module are introduced to achieve better detection accuracy. Finally, to address the boundary discontinuity, the loss function is improved by introducing the IoU constant factor. Considering the problems with the HRSC2016 dataset, we establish a new dataset with more accurate labels and more images and object samples. Through an ablation study, we thoroughly analyze the validity of the proposed rotated RetinaNet, feature alignment module, and improved loss function. The experimental results show that our method is superior to other state-of-the-art methods.

Dual-Resolution and Deformable Multihead Network for Oriented Object Detection in Remote Sensing Images

“…Han et al [28] designed the feature alignment module to refine the anchors. Zhu et al [18] and Chen et al [6] introduced feature alignment modules in RetinaNet to alleviate feature misalignment problems. To accelerate the inference speed and reduce the computational cost, Huang et al [51] devoted efforts to lightweight OOD and proposed LO-Det which achieved real-time detection.…”

P2RNet: Fast Maritime Object Detection From Key Points to Region Proposals in Large-Scale Remote Sensing Images

“…Meanwhile, sub-pixel convolution upsampling is adopted to optimize the feature fusion process. Zhu et al [19] proposed an arbitrary-oriented ship detection network based on RetinaNet, which utilizes rotating anchors and skew NMS to detect rotating objects. In order to improve detection accuracy, It additionally introduces the feature alignment module and the intersection over union (IoU) constant factor.…”

Structures Detection Based on CLSK Model Combined With Shadow Information Using High-Resolution Remote Sensing Images

“…Due to the limited number of targets in a single remote sensing image, which may not effectively evaluate the model's performance, we the test sets for further comparison. We contrasted our network with Rotate Faster RCNN [35], Rotate RetinaNet [36], Rotate FCOS [37], and R3Det [38] networks to assess its performance comprehensively. The parameters for these networks are consistent, including the use of pre-trained weights trained on ImageNet, a batch size of 4, SGD optimizer with a momentum of 0.9, a learning rate of 0.02, and IoU thresholds for positive and negative samples set at 0.3 and 0.7, respectively.…”