一种用于遥感图像目标检测的特征融合检测模型
A Feature Fusion Detection Model for Object Detection in Remote Sensing Images
摘要: 遥感图像和自然图像的差异导致目标检测在遥感图像中效果不佳，因此本文将注意力机制运用到特征提取的过程中，提高特征提取的能力。并使用自注意力机制对所提取各层级特征信息进行融合，基于集成特征做后续的目标检测，提高在目标多尺度问题上的表现。本文在DIOR数据集验证网络模型的可行性，并与当前常见的目标检测模型进行对比验证，取得了67.1%的全类平均精度。实验表明，该模型对比于其他常见目标检测模型在遥感图像上的检测性能有显著地提升。
Abstract: The difference between remote sensing image and natural image leads to poor target detection ef-fect in remote sensing image. Therefore, this paper applies the attention mechanism to the feature extraction process to improve the feature extraction ability. And make the self-attention mechanism to fuse the extracted feature information of each level, and do the follow-up target detection based on the integrated features to improve the performance on the target multi-scale problem. This paper verifies the feasibility of the network model on the DIOR data set, and compares and verifies it with the current common target detection model, and achieves an average accuracy of 67.1% for all classes. Experiments show that compared with other common target detection models, the detection performance of this model on remote sensing images has been significantly improved.
文章引用：刘晓东, 王卓薇, 徐超, 张鳌, 陈海源. 一种用于遥感图像目标检测的特征融合检测模型[J]. 计算机科学与应用, 2021, 11(10): 2538-2545. https://doi.org/10.12677/CSA.2021.1110257

参考文献

[1] Jiao, L., et al. (2019) A Survey of Deep Learning-Based Object Detection. IEEE Access, 7, 128837-128868. [Google Scholar] [CrossRef] 
[2] Ren, S., He, K., Girshick, R. and Sun, J. (2017) Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39, 1137-1149. [Google Scholar] [CrossRef] 
[3] Redmon, J. and Farhadi, A. (2018) YOLOv3: An Incremental Improvement. arXiv:1804.02767 [cs.CV]
[4] Liu, W., Anguelov, D., Erhan, D., et al. (2016) SSD: Single Shot MultiBox Detector. arXiv:1512.02325 [cs.CV] [Google Scholar] [CrossRef] 
[5] Fatima, S.A., Kumar, A., Pratap, A., et al. (2020) Object Recognition and Detection in Remote Sensing Images: A Comparative Study. 2020 International Conference on Artificial Intelligence and Signal Processing (AISP), Amaravati, 10-12 January 2020, 1-5. [Google Scholar] [CrossRef] 
[6] Simonyan, K. and Zisserman, A. (2014) Very Deep Con-volutional Networks for Large-Scale Image Recognition. arXiv:1409.1556 [cs.CV]
[7] He, K., Zhang, X., Ren, S., et al. (2016) Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 27-30 June 2016, 770-778. [Google Scholar] [CrossRef] 
[8] Woo, S., Park, J., Lee, J.Y., et al. (2018) CBAM: Convolutional Block Attention Module. European Conference on Computer Vision, Munich, 8-14 September 2018, 3-19. [Google Scholar] [CrossRef] 
[9] Jie, H., Li, S., Gang, S., et al. (2017) Squeeze-and-Excitation Networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42, 2011-2023. [Google Scholar] [CrossRef] 
[10] Lin, T.Y., Dollar, P., Girshick, R., et al. (2017) Feature Pyra-mid Networks for Object Detection. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Hon-olulu, 21-26 July 2017, 2117-2125. [Google Scholar] [CrossRef] 
[11] Oksuz, K., Cam, B.C., Kalkan, S., et al. (2020) Imbalance Problems in Object Detection: A Review. IEEE Transactions on Pattern Analysis and Machine Intelligence, 43, 3388-3415. [Google Scholar] [CrossRef] 
[12] Wang, X., Girshick, R., Gupta, A., et al. (2017) Non-Local Neural Networks. arXiv:1711.07971 [cs.CV] [Google Scholar] [CrossRef] 
[13] Li, K., Wan, G., Cheng, G., et al. (2019) Object Detection in Opti-cal Remote Sensing Images: A Survey and A New Benchmark. arXiv:1909.00133 [cs.CV] [Google Scholar] [CrossRef]