Investigation of convolutional neural networks for object detection in aerospace images

Vladimir Skripachev, Mikhail Guida, Nikolay Guida, Alexander Zhukov

Abstract


The article discusses current algorithms for solving problems of object recognition in images, their main features and advantages. A brief analysis of existing models of working with images based on convolutional neural networks is carried out. A brief overview of the features of convolutional neural network architectures, quantitative indicators for assessing the quality of their functioning and the types of tasks to be solved, the main features of working with images and the main emerging difficulties are considered, the features of processing aerospace images are highlighted. The problem of object recognition in aerospace images is formulated by adapting existing relevant algorithms and their combinations. The main problems of processing aerospace images and approaches to their solution, the application of established methods of object recognition in conventional images to the problems of object recognition in aerospace images are shown. The analysis of various neural network architectures in the prism of solving object recognition problems in aerospace images is carried out. Conclusions are drawn regarding the most successful combinations of various algorithms in the structure of neural networks when recognizing objects in aerospace images. The main factors that make it difficult to recognize objects in aerospace images and the directions of work to reduce their impact on the accuracy of neural networks when recognizing objects in aerospace images are determined.

Full Text:

PDF (Russian)

References


Adrian Rosebrock. Intersection over Union (IoU) for object detection. https://www.pyimagesearch.com/2016/11/07/intersection-over-union-iou-for-object-detection.

Jonathan Hui. mAP (mean Average Precision) for Object Detection https://jonathan-hui.medium.com/map-mean-average-precision-for-object-detection-45c121a31173.

Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. https://arxiv.org/pdf/1506.01497.pdf.

Wei Liu1, Dragomir Anguelov2, Dumitru Erhan3, Christian Szegedy3, Scott Reed4, Cheng-Yang Fu1, Alexander C. Berg1. SSD: Single Shot MultiBox Detector. https://arxiv.org/pdf/1512.02325.pdf

Joseph Redmon, Santosh Divvala, Ross Girshick, Ali Farhadi. You Only Look Once: Unified, Real-Time Object Detection. https://arxiv.org/pdf/1506.02640.pdf.

Jian Ding, Nan Xue, Gui-Song Xia, Xiang Bai, Wen Yang, Michael Ying Yang, Serge Belongie, Jiebo Luo, Mihai Datcu, Marcello Pelillo, Liangpei Zhang. “Object Detection in Aerial Images: A Large-Scale Benchmark and Challenges”. https://arxiv.org/pdf/2102.12219.pdf.

Z. Liu, H. Wang, L. Weng, and Y. Yang, “Ship rotated bounding box space for ship extraction from high-resolution optical satellite images with complex backgrounds,” IEEE Geosci. Remote Sensing Lett., vol. 13, no. 8, pp. 1074–1078, 2016.

G.-S. Xia, X. Bai, J. Ding, Z. Zhu, S. Belongie, J. Luo, M. Datcu, M. Pelillo, and L. Zhang, “Dota: A large-scale dataset for object detection in aerial images,” in CVPR, 2018.

J. Ding, N. Xue, Y. Long, G.-S. Xia, and Q. Lu, “Learning roi transformer for oriented object detection in aerial images”. https://arxiv.org/pdf/1812.00155.pdf.

J.R.R. Uijlings, K.E.A. van de Sande, T. Gevers2, and A.W.M. Smeulders. Selective Search for Object Recognition. http://www.huppelen.nl/publications/selectiveSearchDraft.pdf.

Vincent Feng. An Overview of ResNet and its Variants. https://towardsdatascience.com/an-overview-of-resnet-and-its-variants-5281e2f56035.

Dong, C., Loy, C.C., He, K., Tang, X.: Learning a deep convolutional network for image super-resolution. In: Computer Vision–ECCV 2014. Springer (2014) 184–199.


Refbacks

  • There are currently no refbacks.


Abava  Кибербезопасность MoNeTec 2024

ISSN: 2307-8162