Machine Learning-Enhanced ORB Matching Using EfficientPS for Error Reduction

[1] In 2019, Viswanathan D G introduced the “Features from Accelerated Segment Test (FAST)” at the 10th Image Analysis for Multimedia Interactive Services Workshop held in London. Search in Google Scholar

[2] In October 2022, Liu B, Yang F, Huang Y, Zhang Y, and Wu G published an article in the “Journal of Applied Science” titled “Single-shot 3D Reconstruction Based on Structured Light Vision Methods.” The article can be found in Volume 12, Issue 20, under article number 10602. Search in Google Scholar

[3] This paper “Strong feature matching of 3D point clouds with asymptotic consistent voting” is co-authored by Quan S., Yin K., Ye K., and Nan K. The paper was published in Volume 22, Issue 20, page 7718 of the journal “Sensor” in October 2022. Search in Google Scholar

[4] Misra, M. K., Rohil, S. M., Moorthi, S., and D. Dhar invented a new system for producing integrated image mosaics nationwide using optical remote sensing images. Their discovery was published in the December 2022 issue of the “Geoinformatics” journal. Search in Google Scholar

[5] In 2018, Zhang B. completed his doctoral thesis at the Graduate School of the Chinese Academy of Sciences. His thesis introduced a technique for using convolutional neural networks for stereo image matching. Search in Google Scholar

[6] Rosen and Delamond presented this research at the European Computer Vision Conference held in Berlin and Heidelberg in 2006, focusing on quickly detecting corners using machine learning techniques. These findings were later published by Springer. Search in Google Scholar

[7] Muja M, Lowe DG. Rapid approximate nearest neighbor search using automatic algorithm configuration. Published at the 2009 International Conference on Computer Vision and Computer Graphics Theory and Applications. Search in Google Scholar

[8] Ding Hui, Li Lihong, and Yuan Gang collaborated on writing an article about computer applications, which was published on page 6 of the 4th issue of the 40th volume in 2020. The title of the article is “Image Alignment Algorithm Combining GMS with VCS+GC-RANSAC”. The DOI is CNKI: SUN: JSJY.0.2020-04-035. Search in Google Scholar

[9] Bian Jie, Lin Wei Yi, and Ji Xin San Lang. GMS: Grid-based motion statistics for fast and reliable feature correspondence. Published at the IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 2017. DOI: 10.1109/CVPR.2017.302. Search in Google Scholar

[10] Fischler (MA) and Bolles (RC) introduced the concept of “random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography” in “Computer Vision and Reading” (1987): 726-740 pages. Search in Google Scholar

[11] Wang, H., Sun, S., Zhou, L., Guo, L., Min, X., Li, C. (2020). Local feature-aware Siamese matching model for vehicle re-identification. Applied Sciences, 10, 2474. https://doi.org/10.3390/app10072474 Search in Google Scholar

[12] Zhu, Zeng, Huang, Liao, Lei, Cai, and Zheng utilized deep learning features to conduct a study on vehicle re-identification in different directions. Their research was published in Volume 21, pages 410-420 of the 2019 IEEE Transactions on Intelligent Transportation Systems journal. Search in Google Scholar

[13] Mo Han, Rocksait and Vallada (2020). “EfficientPS: Efficient Panoramic Segmentation” published on arXiv, preprint number arXiv:2004.02307. Search in Google Scholar

[14] He Kaiming, Georgia Gnnoksalri, Pyotr Dara, and Ross Gillheck (2017) published “Mask R-CNN” in the Proceedings of the IEEE International Conference on Computer Vision (pages 2961-2969). Search in Google Scholar

[15] Zhu, Yu, Hou, Li, Zhang proposed a new visual SLAM system that integrates multiple deep neural networks. Their research was published in Volume 13, page 9630 of the journal “Applied Sciences” in 2023. The article link is https://doi.org/10.3390/app13179630. Search in Google Scholar

[16] Tan, Ming and Li, Qiuwei (2019). “EfficientNet: New Methods for Scalable Convolutional Neural Networks”. arXiv preprint arXiv:1905.11946. Search in Google Scholar

[17] In December 2022, researchers J. Ni, X. Wang, T. Gong, and Y. Xie published a paper titled “Improved Adaptive Monocular Vision Robot ORB-SLAM Method in Dynamic Environments” in Volume 13, Issue 12 of the International Journal of Machine Learning and Control, pages 3821-3836. Search in Google Scholar

[18] In 2017, Wu Y, Ma W, Gong M, and their colleagues introduced a new point matching algorithm for image registration based on rapid sample consistency. This research was published in the “IEEE Geoscience and Remote Sensing Letters” journal, Volume 12, Issue 1, pages 43-47. The publication’s DOI is 10.1109/LGRS.2014.2325970. Search in Google Scholar

[19] Barre, Delepine, Hojeijup, Canipuli (2020) published a research titled “VPS-SLAM: Visual Plane Semantic SLAM for Aerial Robot Systems” in the IEEE Access journal, Volume 8, pages 60704-60718. Search in Google Scholar

[20] Wei, Wang, Li, Liu, Yang, and Liu. Improve vSLAM using semantic information in indoor environments. Applied Science, 2023, 13, 8790. Search in Google Scholar

[21] Zhang Bo and Han Guangliang proposed a method based on Mask R-CNN for removing ORB mismatches. This method was published in Volume 33, Issue 8, on page 7 of the journal “Liquid Crystals and Displays” in 2018. Search in Google Scholar

[22] In their 2022 publication, Dong, Cheng, Peng, and Li introduced a fast semi-direct SLAM algorithm named FSD-SLAM, achieving a speed of 1823-1834. Search in Google Scholar

[23] Wu, Li, Li, Chen, and Li. Visual SLAM feature point tracking method based on multi-constraint changing lighting environment. “Applied Science”, 2023, 13, 7027. Search in Google Scholar

[24] In 2020, Zhang Chunlei and Niu Xingyuan published a research paper integrating the ORB stereo image matching method with YOLO. This research was published in volume 41, issue 1, page 5 of the Journal of Mini-Micro Systems in 2020. Search in Google Scholar

[25] In 2021, a paper titled “Visual SLAM based on Improved ORB Feature Extraction and Affine Transformation” was published in Volume 227, Issue 1 of the journal “Optics”. The authors are Cai Leicai, Ye Yuling, Gao Xiang, Li Zhong, and Zhang Chaoyang. Search in Google Scholar

[26] Wang, Wang, and Tang proposed an improved visual SLAM system that takes into account the reliability of map points in dynamic environments. Their work was published in the Applied Science Journal in 2023. Search in Google Scholar

[27] A fast and efficient accelerator that can extract real-time full HD image features using the ORB algorithm. This research was published in the IEEE Transactions on Very Large Scale Integration (VLSI) Systems in 2020, Volume 28, Issue 2, Pages 565-575. Search in Google Scholar

[28] Redmon, Joseph and others published a report titled “YOLO: Real-Time Object Detection” at the IEEE Conference on Computer Vision and Pattern Recognition in 2016. Search in Google Scholar

[29] Monica Bansal, Munish Kumar, and Manish Kumar conducted a study titled “Comparison and analysis of SIFT, SURF, and ORB feature descriptors for 2D object recognition,” which was published in the journal Multimedia Tools and Applications in 2021, Volume 80(12): 18839-18857. Search in Google Scholar

[30] Rublee, E., Rabaud, V., Konolige, K., et al. (2011). ORB: An efficient alternative to SIFT or SURF. In Proceedings of the International Conference on Computer Vision (pp. 2564-2571). IEEE. Search in Google Scholar

[31] Calonder, M., Lepetit, V., Strecha, C., and Fua, P. (2010). BRIEF: Binary Robust Independent Elementary Features. Presented at the European Conference on Computer Vision in Berlin, Germany. Published by Heidelberg Springer, pages 778-792. Search in Google Scholar

[32] Bay H, Ess A, Tuytelaars T and others. “Speeded-Up Robust Features (SURF).” Computer Vision and Image Understanding, 2008, Vol. 110, No. 3, pp. 346-359. Search in Google Scholar

[33] Lin WYD, Cheng MM, Lu J, and their colleagues presented a paper titled “Bidirectional Function of Global Action Modeling” at the 2014 European Conference on Computer Vision. The paper can be found at DOI: 10.1007/978-3-319-10593-2_23. Search in Google Scholar

[34] Authors such as K He, G Gkioxari, and P Dollár. “Mask R-CNN” was published in 2017 in IEEE Transactions on Pattern Analysis and Machine Intelligence. DOI: 10.1109/TPAMI.2018.2844175. Search in Google Scholar