Hydrological monitoring in small catchments: the MagicHydroBox

The monitoring of small headwater catchment represents a major issue in hydrology, especially in remote areas, where gathering real-time hydrological data is often prohibitive due to the limited availability of power and connectivity. However, recent advances in non-contact computer vision and informatic technology offer an opportunity to fill such technical gap. In this regard, we designed and developed the MagicHydroBox prototype (MHB), and all-in-one camera and processing unit system aimed at monitoring the water level in small headwater rivers. The tasks performed by MHB include image collection, image processing, storage and transmission of the processed data. Since the MHB is equipped with NIR (NearInfrared) leds and camera, the image collection can be carried out both during the day and the night period. The image processing takes place directly in the MHB, to guarantee the onsite analysis, it is based on the Otsu’s segmentation method to identify a properly placed target within the images, and results in the direct estimation of water depth. Finally, we built in the MHB the possibility to transmit the processed data both through Gprs (mobile data) and LoRaWan (a long-range, low-power system). The MHB is also equipped with a GUI that allows the user to set and calibrate the instrument. We carried out preliminary field tests to evaluate the effectiveness of the MHB in providing an accurate measure of the target and transmitting the processed data. The preliminary results highlight the potential of the MHB to estimate the water level, especially in NIR images, and to provide a real-time hydrological monitoring where Internet signal is available. The main innovation of the MHB is represented by the fact that it automated a series of tasks that were instead manually performed in previous works. The concentration of all the necessary tasks within the MHB simplify the data acquisition, the processing and the management providing an useful tool where frequent maintenance or monitoring surveys are not possible. Moreover, the MHB is promising for future implementation of algorithms to measure surface velocimetry and discharge.