| Title | Water Level Estimation Based on Image of Staff Gauge in Smart City |
|---|---|
| ID_Doc | 39500 |
| Authors | Xu, ZK; Feng, J; Zhang, ZZ; Duan, CF |
| Title | Water Level Estimation Based on Image of Staff Gauge in Smart City |
| Year | 2018 |
| Published | |
| Abstract | In the automatic measurement of water level, the hydrological satiation also uses a traditional staff gauge to calibrate the water depth measured by the sensor. And in the most of water-covered areas, such as underpass overpasses, low-lying roads and tunnels, staff gauges have been installed to monitor road water. Therefore, the automatic water level measurement method based on staff gauge image, has drawn more and more attention. So we propose a method for the water level estimation based on images of staff gauge in this paper. First, we proposed a two-step method for segmenting the staff gauge. We implemented the initial positioning by HSV color space and morphological processing, and then we defined a component map to achieve precise positioning. Next, we split the characters on the staff gauge image and recognized those characters through a Convolutional Neural Network. Finally, we measured the water level by using a quadratic function to determine the mapping relationship between pixels and the metric values. Compared with the traditional method of extracting water line, the two-step positioning method improves the water level detection accuracy. The measurement result in a real world image of a pond shows more accurate detection with the algorithm and the feasibility of applying it to road water monitoring. |
No similar articles found.