|
|
基于YOLOv5算法的水位智能监测系统
|
Abstract:
近年来，图像识别技术快速发展，为了实现水位数据前端处理，本论文提出了将YOLOv5算法用于前端处理水位监测数据的方案。本方案与传统方案最大的不同之处在于数据处理方式，本方案直接将YOLOv5算法嵌入集成单片机中，在前端进行数据转换，最后将结果显示在终端，传统方案需要将数据传输回服务器中。论文实验数据的采集，考虑了不同测量地点和光照角度等因素，具有真实性。最后将YOLOv5算法的实验数据与YOLOv4、hog + svm和Faster RCNN算法进行对比，并以F1作为比较标准。结果显示：在验证集下，YOLOv5算法的mAP值和准确率等指标均表现优异，F1值高于Faster RCNN将近4%，高于HOG + SVM将近12%，在GPU和CPU环境下，单张图片检测时间分别为15 ms、125 ms，说明YOLOv5算法在各种复杂环境下的均能适应图像识别任务，具有高识别率和鲁棒性。
In recent years, image recognition technology has developed rapidly, in order to achieve front-end processing of water level data, this paper mainly proposes a scheme of using YOLOv5 algorithm for front-end processing of water level monitoring data. The biggest difference between this scheme and traditional schemes lies in the data processing method. This scheme directly embeds the YOLOv5 algorithm into the integrated microcontroller, performs data conversion in the front-end, and finally displays the results in the terminal. Traditional schemes require data transmission back to the server. The collection of experimental data in the paper takes into account factors such as different measurement locations and lighting angles, and has authenticity. Finally, compare the ex-perimental data of YOLOv5 algorithm with YOLOv4, hog + SVM, and Faster RCNN algorithms, and use F1 as the comparison standard. The results show that under the validation set, the YOLOv5 algorithm performs excellently in terms of mAP value and accuracy, with an F1 value nearly 4% higher than Faster RCNN and nearly 12% higher than HOG + SVM. In GPU and CPU environments, the single image detection time is 15 ms and 125 ms, respectively. This indicates that the YOLOv5 algorithm can adapt to image recognition tasks in various complex environments, with high recognition rate and robustness.
| [1] | 孙英豪, 丁勇, 李登华, 谢东辉. 基于图像识别的无标尺水位测量技术研究[J/OL]. 水利水运工程学报: 1-9.
http://kns.cnki.net/kcms/detail/32.1613.TV.20230227.1329.004.html, 2023-02-28. |
| [2] | 杨宇, 刘宇红, 彭燕, 孙雨琛, 张荣芬. 基于机器视觉的目标识别追踪算法及系统设计[J]. 传感器与微系统, 2020, 39(4): 92-95+98. |
| [3] | 张帆, 靳晓妍. 基于视频图像的嵌入式水位监测方法[J]. 中国测试, 2022, 48(12): 140-145. |
| [4] | 李涛, 徐高, 梁思涵, 李英睿, 王敏, 李冰. 人工智能图像识别在水利行业的应用进展[J]. 人民黄河, 2022, 44(11): 163-168. |
| [5] | 傅启凡, 路茗, 张质懿, 纪立, 丁华泽. 基于语义分割的水位监测方法研究[J]. 激光与光电子学进展, 2022, 59(4): 89-100. |
| [6] | 周婷, 汪炎, 邹俊, 李辰, 崔玉环, 王笑宇, 谢传流, 夏萍. 基于PCA和SVM的遥感影像水体提取方法及验证[J]. 水资源保护, 2023, 39(2): 180-189. |
| [7] | 杨有为, 周刚. 面向自然场景文本检测的改进NMS算法[J]. 计算机工程与应用, 2022, 58(1): 204-208. |
| [8] | Bharati, P. and Pramanik, A. (2020) Deep Learning Tech-niques—R-CNN to Mask R-CNN: A Survey. In: Das, A., Nayak, J., Naik, B., Pati, S. and Pelusi, D., Eds., Computa-tional Intelligence in Pattern Recognition. Advances in Intelligent Systems and Computing, Vol. 999, Springer, Singa-pore, 657-668. https://doi.org/10.1007/978-981-13-9042-5_56 |
| [9] | Mikkelsen, L., Moesgaard, K., Hegnauer, M. and Lopez, A.D. (2020). ANACONDA: A New Tool to Improve Mortality and Cause of Death Data. BMC Medicine, 18, Article No. 61. https://doi.org/10.1186/s12916-020-01521-0 |
| [10] | Men, Q. and Shum, H.P.H. (2022) PyTorch-Based Implementation of Label-Aware Graph Representation for Multi-Class Trajectory Prediction. Software Impacts, 11, Article ID: 100201. https://doi.org/10.1016/j.simpa.2021.100201 |
| [11] | Li, F., Wang, Y., Jiang, J., et al. (2023) Heterogeneous Acceleration Algorithms for Shallow Cumulus Convection Scheme over GPU Clusters. Future Generation Computer Systems, 146, 166-177.
https://doi.org/10.1016/j.future.2023.04.021 |
| [12] | 江铃燚, 郑艺峰, 陈澈, 李国和, 张文杰. 有监督深度学习的优化方法研究综述[J]. 中国图象图形学报, 2023, 28(4): 963-983. |
| [13] | 刘昱, 于坤霞, 李鹏, 李占斌, 张晓明. 黄河中游典型流域水文统计模型精度集合评价[J]. 人民黄河, 2023, 45(4): 20-27. |
| [14] | 周熙然, 李德仁, 薛勇, 汪云甲, 邵振峰. 地图图像智能识别与理解: 特征、方法与展望[J]. 武汉大学学报(信息科学版), 2022, 47(5): 641-650. |
| [15] | Wu, S., Yang, J., Wang, X. and Li, X. (2022) IoU-Balanced Loss Functions for Single-Stage Object De-tection. Pattern Recognition Letters, 156, 96-103. https://doi.org/10.1016/j.patrec.2022.01.021 |
| [16] | Cao, Y., Chen, Z., Wen, T., et al. (2022) Rail Fastener Detection of Heavy Railway Based on Deep Learning. High-Speed Railway, 1, 63-69. https://doi.org/10.1016/j.hspr.2022.11.001 |
| [17] | Ouf, N.S. (2023) Leguminous Seeds Detection Based on Convolutional Neural Networks: Comparison of Faster R-CNN and YOLOv4 on a Small Custom Dataset. Artificial In-telligence in Agriculture, 8, 30-45.
https://doi.org/10.1016/j.aiia.2023.03.002 |
| [18] | Yuan, J., Zheng, X., Peng, L., et al. (2023) Identification Method of Typical Defects in Transmission Lines Based on YOLOv5 Object Detection Algorithm. Energy Reports, 9, 323-332. https://doi.org/10.1016/j.egyr.2023.04.078 |
| [19] | Sharma, A.K., Nandal, A., Dhaka, A., et al. (2023) HOG Trans-formation Based Feature Extraction Framework in Modified Resnet50 Model for Brain Tumor Detection. Biomedical Signal Processing and Control, 84, Article ID: 104737. https://doi.org/10.1016/j.bspc.2023.104737 |
| [20] | Bandong, S., Nazaruddin, Y.Y. and Joelianto, E. (2023) Faster RCNN Mixed-Integer Optimization with Weighted Cost Function for Container Detection in Port Automation. Heliyon, 9, E13213.
https://doi.org/10.1016/j.heliyon.2023.e13213 |
