Research on Water-Level Recognition Method Based on Image Processing and Convolutional Neural Networks

Abstract

Water level dynamics in catchment-scale rivers is an important factor for surface water studies. Manual measurement is highly accurate but inefficient. Using automatic water level sensors has disadvantages such as high cost and difficult maintenance. In this study, a water level recognition method based on digital image processing technology and CNN is proposed. For achieving batch segmentation of source images, the coordinates of the water ruler region in the source image and characters’ region and the scale lines’ region on the ruler are obtained by using image processing algorithms such as grayscale processing, edge detection, and the tilt correction method based on Hough-transform and morphological operations. The CNN is then used to identify the value of digital characters. Finally, the water level value is calculated according to the mathematical relationship between the number of scale lines detected by pixel traversal in the binarized image and the value of digital characters. This method is used to identify the water levels of the water ruler images collected in the Hulu watershed of the Qilian Mountains in Northwest China. The results show that the accuracy compared with the actual measured water level reached 94.6% and improved nearly 24% compared to the template matching algorithm. With high accuracy, low cost, and easy deployment and maintenance, this method can be applied to water level monitoring in mountainous rivers, providing an effective tool for watershed hydrology research and water resources management.

Evaluation of deep learning computer vision for water level measurements in rivers

“…The existing literature on the application of deep learning computer vision to water level analysis has predominantly focused on comparing the differences among various deep learning methods [ 23 , 27 , 43 , 44 ] or comparing the water level analyzed by deep learning with the measured water level [ 17 , 45 , 46 ]. However, traditional image processing methods have not been discussed or compared, and only a few studies have examined the effect of image datasets on the analysis results during deep learning network training.…”

Rulers2023: An Annotated Dataset of Synthetic and Real Images for Ruler Detection Using Deep Learning

“…As the results indicate, these techniques offer a low-cost solution to monitor urban flooding, especially in areas susceptible to deep-water accumulation. Another notable application is in monitoring water levels in catchment-scale rivers [15]. Although accurate, traditional manual measurements are inefficient, and automatic sensors come with their own challenges.…”

“…One of the seemingly simple but critical objects that demand accurate detection is the measurement ruler. Algorithms for ruler detection, recognition, and interpretation are central to the accuracy and efficacy of systems, facilitating automated, high-precision measurements in fields such as aquaculture [10][11][12][13], environmental monitoring [14][15][16][17][18][19][20], medical diagnostics [21][22][23][24][25][26][27][28][29][30][31][32], forensics [33,34], industry [35][36][37][38], museums [39]. However, the task of detecting rulers in images presents challenges that have not been extensively explored in current research (Figure 1).…”

Automatic Monitoring Alarm Method of Dammed Lake Based on Hybrid Segmentation Algorithm

“…In addition, when the image template is extracted from the marker in advance, the part of the template that can be matched by the multi-template matching technology is the area that has not been covered by the water, allowing the variation in water level to be obtained. [ 22 , 23 ]. The manual placement of water gauges is highly limited because it is often inaccessible at many monitoring sites.…”