Experimental Evaluation of the Kinect Sensor for Hydro-Geomorphological Measurements: Limitations, and Future Opportunities
摘要
This paper presents experimental observations and practical insights into the application of the Microsoft Kinect sensor for hydro-geomorphological research, based on the authors’ investigations of riverbank erosion monitoring, scour hole characterization, and gravel-bed morphology analysis. The Kinect, a low-cost infrared-based 3D imaging device, offers high spatial and temporal resolution, making it a promising low-cost complementary tool for geomorphological measurements alongside to traditional methods like terrestrial laser scanning and photogrammetry. Experimental results demonstrate the device’s high accuracy in controlled environments, with volumetric measurement errors as low as 0.18% for hemispherical and 0.5% for conical volumes. In seepage erosion studies, Kinect effectively captured cavity dynamics under varying hydraulic heads. In scour hole experiments, measurement errors ranged from 1.1% to 11.22%, with refraction corrections applied in underwater conditions. Gravel bed characterization resulted in an absolute mean error of 2.4% under controlled lighting, though significant limitations emerged under direct sunlight interference. Future research should aim to mitigate environmental limitations through improved infrared management, integrate Kinect data with other imaging systems for enhanced accuracy, and develop standardized protocols. Expanded applications under varied hydraulic conditions and detailed characterization of natural gravel bed morphology are also recommended to fully realize the Kinect’s potential in hydro-geomorphological studies.