Iran-Water Resources Research

Iran-Water Resources Research

Comparing the Efficacy of Deep Learning Models and Convolutional Neural Network for Water Level Estimation Using Smartphone, and Image Processing

Document Type : Original Article

Authors
Erfan Abdi 1
Rasoul Jani 2
Sabere Darbandi 3
1 M.Sc. Student, Department of Water Engineering, Faculty of Agriculture, University of Tabriz, Iran.
2 Assistant Professor, Department of Civil Engineering, Tabriz Branch, Islamic Azad University, Tabriz, Iran.
3 Associate Professor, Department of Water Engineering, University of Tabriz, Tabriz, Iran.
10.22034/iwrr.2024.445181.2744
Abstract
Water level estimation is essential for understanding and managing water resources, predicting and mitigating flood effects, and water resource management decisions but it is often hindered by limited resources, capitals and time. This study investigates the performance of a non-contact, image-based water level measurement system using advances in smartphone imaging technology. For this purpose, a smartphone equipped with a camera was used to record the desired images. The images were then processed by image processing to identify and measure the water level. The core of this study involved the development and comparison of two computational models: deep learning (DL) and convolutional neural networks (CNN). These models were used to estimate the water level based on the processed image data. The results showed a different degree of accuracy among the models; the CNN model showed better performance than the DL, which had the lowest root mean square error of 24.36 mm. In contrast, the DL model with the root mean square error of 28.39 mm showed that the relative effectiveness of the convolutional neural network in this study is better than deep learning in terms of accuracy and high reliability. The study showed that it is possible to monitor and control the water level in remote and difficult places completely automatically without the need for relevant personnel.
Keywords
Non-Contact Measurement
Hydrological Monitoring
Image Analysis
Convolutional Neural Networks
Computational Hydrology

Subjects

Acharya U R, Fujita H, Lih OS, Hagiwara Y, Tan J H, and Adam M (2017) Automated detection of arrhythmias using different intervals of tachycardia ECG segments with convolutional neural network. Information Sciences 405:81-90
 Birgand F, Chapman K, Hazra A, Gilmore T, Etheridge R, and Staicu A M (2022) Field performance of the GaugeCam image-based water level measurement system.  PLoS Water 1(7):p.e0000032
Chaudhary P, D'Aronco S, Moy de Vitry M, Leitão J P, & Wegner J D (2019) Flood-water level estimation from social media images. ISPRS Annals of the Photogrammetry. Remote Sensing and Spatial Information Sciences 4(2/W5):5-12
Dou G, Chen R, Han C, Liu Z, and Liu J (2022) Research on water-level recognition method based on image processing and convolutional neural networks. Water 14(12):1890
Dumoulin V and Visin F (2016) A guide to convolution arithmetic for deep learning. arXiv preprint arXiv:1603.07285
Deng J, Dong W, Socher R, Li LJ, Li K, and Fei-Fei L (2009) Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, pp.248-255
Elias M and Maas H G (2022) Measuring water levels by handheld smartphones: A contribution to exploit crowdsourcing in the spatio-temporal densification of water gauging networks. The International Hydrographic Review (27):pp.9-22:(27)28
Fleury G R D O, do Nascimento D V, Galvão Filho A R, Ribeiro F D S L, de Carvalho R V, and Coelho C J (2020) Image-based river water level estimation for redundancy information using deep neural network. Energies 13(24):p.6706
Gao A, Wu S, Wang F, Wu X, Xu P, Yu L, and Zhu S (2019) A newly developed unmanned aerial vehicle (UAV) imagery based technology for field measurement of water level. Water 11(1):124
Gu M, Su B, Wang M, and Wang Z (2019) Survey on decolorization methods. Journal of Applied Computing Research 36:1286-1292
Hies T, Parasuraman S, Wang Y A D O N G, Duester R U D O L P H, Eikaas H, and Tan K M (2012) Enhanced water-level detection by image processing. In 10th International Conference on Hydroinformatics (Vol. 1).  Hamburg, Germany
Krizhevsky A, Sutskever I, and Hinton G E (2017) ImageNet classification with deep convolutional neural networks. Communications of the ACM 60(6):pp.84-90
Kharade A, Gendle M, and Lodha T, (2017) Water level measurement and detection of flow direction using image processing. International Journal of Innovations in Engineering Research and Technology: pp.1-4:(52)17
Kuo L C, and Tai C C (2022) Robust image-based water-level estimation using single-camera monitoring. IEEE Transactions on Instrumentation and Measurement 71:1-11
Kumar T and Verma K (2010) A Theory Based on Conversion of RGB image to gray image. International Journal of Computer Applications 7(2):7-10
Kim D, Alber M, Kwok M W, MitroviĆ J, Ramirez-Atencia C, PÉrez J A R, and Zille H (2022) Clarifying assumptions about artificial intelligence before revolutionising patent law. GRUR International 71(4):295-321
Kumar T and Verma K (2010) A theory based on conversion of RGB image to Gray image. International Journal of Computer Applications 7(2):7-10
LeCun Y, Bengio Y, and Hinton G (2015) Deep learning. Nature 521(7553):436-444
Li E, Xia J, Du P, Lin C, and Samat A (2017) Integrating multilayer features of convolutional neural networks for remote sensing scene classification. IEEE Transactions on Geoscience and Remote Sensing 55(10):5653-5665
Muhadi N A, Abdullah A F, Bejo S K, Mahadi M R, and Mijic A (2021) Deep learning semantic segmentation for water level estimation using surveillance camera. Applied Sciences 11(20):9691
Noto S, Tauro F, Petroselli A, Apollonio C, Botter G, and Grimaldi S (2022) Low-cost stage-camera system for continuous water-level monitoring in ephemeral streams. Hydrological Sciences Journal 67(9):1439-1448
Nicholaus I T, Lee J S, and Kang D K (2022) One-class convolutional neural networks for water-level anomaly detection. Sensors 22(22):8764
Ortigossa E S, Dias F, Ueyama J, and Nonato L G (2015) Using digital image processing to estimate the depth of urban streams. In Proceedings of the Workshop of Undergraduate Works in Conjunction with Conference on Graphics, Patterns and Images (SIBGRAPI), Bahia, Brazil: pp. 26-29
Qiao G, Yang M, and Wang H (2022) A water level measurement approach based on YOlOv5s. Sensors 22(10):3714
Rusk N (2016) Deep learning. Nature Methods 13(1):35
Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, and Rabinovich A (2015) Going deeper with convolutions. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition: pp. 1-9
Sauvola J and Pietikäinen M (2000) Adaptive document image binarization. Pattern Recognition 33(2):225-236
Sabbatini L, Palma L, Belli A, Sini F, and Pierleoni P (2021) A computer vision system for staff gauge in river flood monitoring. Inventions 6(4):79
Vandaele R, Dance S L, and Ojha V (2021) Deep learning for automated river-level monitoring through river-camera images: an approach based on water segmentation and transfer learning. Hydrology and Earth System Sciences 25(8):4435-4453
Vandaele R, Dance S L, and Ojha V (2023) Calibrated river-level estimation from river cameras using convolutional neural networks. Environmental Data Science 2:1-19:(11)2
Zhang Z, Zhou Y, Liu H, and Gao H (2019) In-situ water level measurement using NIR-imaging video camera. Flow Measurement and Instrumentation 67:95-106
 

  • Receive Date 21 February 2024
  • Revise Date 24 April 2024
  • Accept Date 22 June 2024