Our work facilitates the application of deep optical flow on RSV estimation and provides other related researches with optical flow datasets for fine-tuning. Experiment results demonstrate that the generated virtual river datasets effectively improve the generalization of the model and MRAFT remarkably alleviates the mismatch. Specifically, we introduced a method for generating relative optical flow datasets and proposed MRAFT, a deep optical flow estimation model combining with correlation volume modulation. This work proposed a method for accurately and robustly RSV estimation under velocity range of 0-6.0 m/s. Besides, high similarity of river surfaces often leads to ambiguous correlation volume extracted by deep optical flow estimation model, which can cause mismatch. Models obtained by such approaches suffer from limited generalization because of domain drift. However, these methods often use irrelevant datasets for training due to the immeasurability of optical flow. As a powerful approach for estimate optical flow accurately and efficiently, deep optical flow estimation is utilized in OFV-based methods as well. Among these methods, OFV-based methods have drawn great attention due to its high field resolution and low requirement of tracers. Many image-based velocimetry methods, such as large-scale particle image velocimetry (LSPIV), space-time image velocimetry (STIV) and optical flow velocimetry (OFV), have been proposed to estimate river surface velocity (RSV) efficiently and precisely. Another method, based on the analysis of real trajectory of the boat (obtained from topographic measurement or GPS positioning) compared with the ADCP computed trajectory, is under study. This method, systematically applied to ADCP discharge measurements obtained at Ă“bidos hydrometric station, allowed all measured discharges to be corrected, especially for 19 floods. A correction method was developed on the basis of this correlation. This has allowed quantification of river bed load speed, or bottom displacement speed. during low flow period, this return position error is weak (less than 50 m). When there is no bottom displacement, i.e. It was possible to establish a correlation between the water velocity close to the river bottom and the error between real position and position computed by ADCP when the boat returns to its starting point after a two-way crossing of the river. This error leads to an underestimation of discharge value. Implementation of modern discharge measurement techniques using ultrasonic devices (ADCP), give evidence of a systematic error linked to the displacement of the river bottom due to high water velocity close to the bottom. Since 1995, hydrologists of the HiBAm (Hydrology and Geochemistry of the Amazon Basin) Research Program carried out several hundred discharge measurements in the Amazon basin.
0 Comments
Leave a Reply. |