Single Shot MultiBox Detector-based Feature Fusion Model for Building Object Detection

Xiaoyan  Zheng; Xiaoli  Zhao

doi:10.6180/jase.202502_28(2).0017

Single Shot MultiBox Detector-based Feature Fusion Model for Building Object Detection

Xiaoyan Zheng and Xiaoli ZhaoThis email address is being protected from spambots. You need JavaScript enabled to view it.

Department of Civil Engineering and Architecture, Zhengzhou University of Science and Technology, Zhengzhou, 450064, China

Received: February 26, 2024
Accepted: March 25, 2024
Publication Date: April 29, 2024

Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.

Download Citation: ||https://doi.org/10.6180/jase.202502_28(2).0017

Aiming at the problems of low detection accuracy and large size of existing building detection models, which lead to the imbalance between remote sensing image detection speed and accuracy and are not conducive to later deployment, a building detection method based on single shot multiBox detector (SSD)-based feature fusion model is proposed. In this method, the feature extraction module extracts the features from the input image, and the discrimination features and interference features are decomposed by the feature decomposition module. Finally, the identification features are input into the multi-scale detection module for target detection. The interference features removed after feature decomposition are unfavorable to target detection, including complex background clutter, while the retained identification features are favorable to target detection, including targets of interest, thus effectively reducing false alarms and missing alarms and improving the detection performance of building targets. Experiments show that the interaction ratio, accuracy and total accuracy of proposed method on WHU data set reach 94.2%, 97.0% and 98.9% respectively, showing good effectiveness without significantly increasing parameters.

Keywords: building detection; single shot multiBox detector; feature fusion; multi-scale detection

[1] P. Wang, B. Bayram, and E. Sertel, (2022) “A comprehensive review on deep learning based remote sensing image super-resolution methods" Earth-Science Reviews 232: 104110.
[2] S. Yin, (2023) “Object Detection Based on Deep Learning: A Brief Review" IJLAI Transactions on Science and Engineering 1(02): 1–6.
[3] Y. Ye, T. Tang, B. Zhu, C. Yang, B. Li, and S. Hao, (2022) “A multiscale framework with unsupervised learning for remote sensing image registration" IEEE Transactions on Geoscience and Remote Sensing 60: 1– 15.
[4] R. Girshick, J. Donahue, T. Darrell, and J. Malik, (2015) “Region-based convolutional networks for accurate object detection and segmentation" IEEE transactions on pattern analysis and machine intelligence 38(1): 142–158.
[5] R. Girshick. “Fast r-cnn”. In: Proceedings of the IEEE international conference on computer vision. 2015, 1440–1448.
[6] S. Ren, K. He, R. Girshick, and J. Sun, (2016) “Faster R-CNN: Towards real-time object detection with region proposal networks" IEEE transactions on pattern analysis and machine intelligence 39(6): 1137–1149.
[7] W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y. Fu, and A. C. Berg. “Ssd: Single shot multibox detector”. In: Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14. Springer. 2016, 21–37.
[8] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi. “You only look once: Unified, real-time object detection”. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016, 779–788.
[9] B. Liu, J. Luo, and H. Huang, (2020) “Toward automatic quantification of knee osteoarthritis severity using improved Faster R-CNN" International journal of computer assisted radiology and surgery 15: 457–466.
[10] L. Shen, H. Tao, Y. Ni, Y. Wang, and V. Stojanovic, (2023) “Improved YOLOv3 model with feature map cropping for multi-scale road object detection" Measurement Science and Technology 34(4): 045406.
[11] J. Jang, D. Van, H. Jang, D. H. Baik, S. Duk Yoo, J. Park, S. Mhin, J. Mazumder, and S. H. Lee, (2020) “Residual neural network-based fully convolutional network for microstructure segmentation" Science and Technology of Welding and Joining 25(4): 282–289.
[12] Y. Liu, B. Fan, L. Wang, J. Bai, S. Xiang, and C. Pan, (2018) “Semantic labeling in very high resolution images via a self-cascaded convolutional neural network" ISPRS journal of photogrammetry and remote sensing 145: 78–95.
[13] J. Ding, J. Zhang, Z. Zhan, X. Tang, and X. Wang, (2022) “A precision efficient method for collapsed building detection in post-earthquake UAV images based on the improved NMS algorithm and Faster R-CNN" Remote Sensing 14(3): 663.
[14] T. Bai, Y. Pang, J. Wang, K. Han, J. Luo, H. Wang, J. Lin, J. Wu, and H. Zhang, (2020) “An optimized faster R-CNN method based on DRNet and RoI align for building detection in remote sensing images" Remote Sensing 12(5): 762.
[15] G. Hua, M. Liao, S. Tian, Y. Zhang, and W. Zou, (2023) “Multiple relational learning network for joint referring expression comprehension and segmentation" IEEE Transactions on Multimedia:
[16] W. Gaihua, L. Jinheng, C. Lei, D. Yingying, and Z. Tianlun, (2022) “Instance segmentation convolutional neural network based on multi-scale attention mechanism" Plos one 17(1): e0263134.
[17] F. N. Iandola, S. Han, M. W. Moskewicz, K. Ashraf, W. J. Dally, and K. Keutzer, (2016) “SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size" arXiv preprint arXiv:1602.07360:
[18] D. Sinha and M. El-Sharkawy. “Thin mobilenet: An enhanced mobilenet architecture”. In: 2019 IEEE 10th annual ubiquitous computing, electronics & mobile communication conference (UEMCON). IEEE. 2019, 0280–0285.
[19] M. Tan, R. Pang, and Q. V. Le. “Efficientdet: Scalable and efficient object detection”. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020, 10781–10790.
[20] K. Han, Y. Wang, Q. Tian, J. Guo, C. Xu, and C. Xu. “Ghostnet: More features from cheap operations”. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020, 1580–1589.
[21] S. Yin, L. Wang, Q. Wang, M. Ivanovi´c, and J. Yang, (2023) “M2F2-RCNN: Multi-functional faster RCNN based on multi-scale feature fusion for region search in remote sensing images" Computer Science and Information Systems (00): 54–54.
[22] M. Luo, S. Ji, and S. Wei, (2023) “A diverse large-scale building dataset and a novel plug-and-play domain generalization method for building extraction" IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing:
[23] H. Cao, Y. Wang, J. Chen, D. Jiang, X. Zhang, Q. Tian, and M. Wang. “Swin-unet: Unet-like pure transformer for medical image segmentation”. In: European conference on computer vision. Springer. 2022, 205–218.
[24] G. Wu, X. Shao, Z. Guo, Q. Chen, W. Yuan, X. Shi, Y. Xu, and R. Shibasaki, (2018) “Automatic building segmentation of aerial imagery using multi-constraint fully convolutional networks" Remote Sensing 10(3): 407.
[25] Z. Zhou, M. M. R. Siddiquee, N. Tajbakhsh, and J. Liang, (2019) “Unet++: Redesigning skip connections to exploit multiscale features in image segmentation" IEEE transactions on medical imaging 39(6): 1856–1867.
[26] L. Teng, Y. Qiao, M. Shafiq, G. Srivastava, A. R. Javed, T. R. Gadekallu, and S. Yin, (2023) “FLPK-BiSeNet: Federated learning based on priori knowledge and bilateral segmentation network for image edge extraction" IEEE Transactions on Network and Service Management: