Linear erosion channel (LEC) devastates arable land and significantly contributes to soil loss in agricultural watersheds. In the presence of a less- or non-erodible layer, channel widening governs the erosion process once the channel bed incises to this layer, accompanied by failure block generation and transport. Current knowledge on channel widening, however, is limited due to the lack of robust and efficient methods to capture the rapid sidewall expansion process. Laboratory experiments were designed to simulate the channel widening process with an initial channel width of 10 cm. Two packed soil beds with a non-erodible layer and two slope gradients (5 % and 11 %) were subjected to the inflow rate of 0.67 L/s. Images were captured by mounted digital cameras and automatically transformed into orthophotos. Channel edges and failure blocks were automatically detected by deep learning algorithm in a newly developed Channel-DeepLab network model based upon DeepLabv3+ platform. The procedure includes learning samples labelling, data augmentation, model construction, training, and validation. Sediment discharge and changes in channel width, geometry of channel edges, and failure blocks were measured. The results indicate that initial period is critical for erosion prediction and remediation due to its small sidewall failure interval, high channel expansion rate and sediment discharge. Channel surface area has great potential on accumulated sediment discharge prediction. The slope section that witnessed the fastest channel widening rate migrated downwards when slope gradient increased from 5 % to 11 %. The total number and area of the failure blocks increased with time, while the collapse frequency of the sidewalls decreased. Upstream reach experienced the highest sidewall collapse frequency and rate of disaggregation and transport, while the downstream reach experienced the highest total number of failure blocks. A time lag was found between sidewall collapse and sediment discharge, which increased as time progressed, attributing to decreased runoff erosivity as the flow velocity decreased. Results of this study will provide methodological support for channel sidewall and streambank retreat monitoring, realizing the automatic detection of channel edges and efficient output of rapid sidewall expansion process with high temporal and spatial precision. Future work can be focused on broadening the applicability of the Channel-DeepLab network model and quantifying the delayed response process between sidewall failure and sediment discharge.

建立了基于运动结构恢复方法和特征点匹配测速技术的无人机巡航测速系统,实现了对河流大范围表面流场的长距离巡航监测.首先建立地面控制点,并向水中投入示踪物体;利用无人机巡航俯视拍摄河流表面,连续釆集照片;采用运动恢复结构方法估算各时刻相机的姿态参数,建立目标区域正射投影网格,转换得到正射影图像;再利用加速鲁棒特征算法识别和匹配连续粒子图像中的特征点,得到特征点坐标和位移;进行尺度转化,最终得到流场.该方法成功应用于瑞士苏黎世利马特河表面流场测量,结果表明该方法得到的正射投影图像质量高,流场计算结果合理,与PIV方法相比误差为3.4％,计算速度提升8倍;计算结果与水文站断面实测值吻合良好.

A new airborne river surface flow measurement technique is presented, called Airborne Feature Matching Velocimetry (AFMV). It uses matchings of characteristic image features for orthorectification and velocimetry. Riparian matching points with an arbitrarily chosen base image serve to find individual projective transformation matrices to stabilize airborne video recordings. Transformed matching points’ distances of feature shifts between subsequent video frames lead to surface velocity vectors. To test this approach, a riverine moving water surface was recorded by an airborne video camera. Results are compared to (i) image frames rectified by 3D photogrammetry and (ii) related velocimetry results obtained by Particle Image Velocimetry (PIV). Generally, both time averaged and instantaneous surface velocities obtained by AFMV are shown to be of almost equal quality to the PIV approach. AFMV gives slightly less spatially-dense results in poorly textured areas, but clearly outperforms the 3D photogrammetry reference method in relation to computational power. Thus, the new method bears the potential to provide almost real-time instantaneous airborne river surface flow measurements.

多沙河流上修建的水库必然要面对泥沙淤积问题,合理科学的库容恢复和保持措施对于水库综合效益的正常发挥具有重要意义.排沙廊道是一种一次建设、可持续使用的排沙设施,但传统排沙廊道中各进口为顺序排列,不同部位进口抽沙能力不均,严重影响运行排沙效果.等阻力树权型冲沙管道是近期提出的一种新型排沙方式,其各级管道采用树权型布置,多口汇流的旋转流动可消除不同角度水流加速度影响,按每一个进口管至总出口的流动阻力相等为原则设计;在进口管顶部设置倒扣铁锅型防沙罩,防止进口管被落淤的泥沙所阻塞.本文在水槽中开展了树权型冲沙管道试验,同时开展了相同条件下的进沙口顺序排列的对比试验,结果表明:当进沙口按传统廊道多口顺序排列布置时,位于远端的一半出口将被完全淤堵,难以达到预期的排沙效果;而树权型冲沙管有效克服了传统廊道不同进沙口抽沙能力不均的缺陷,实现了对所覆盖区域的均匀排沙,正常情况下未发生淤堵;当高含沙水流集中流向某一进口时,会造成该进口淤堵,但由于不同进沙口独立运行,所淤堵进口基本不会对其他进口的正常运行造成影响.本文试验结果表明,等阻力树权型冲沙管道的稳定排沙浓度可达400～700 kg/m3,远高于顺序排列方式的100 kg/m3;冲刷漏斗体积为多口顺序排列的1.5倍.

为提升室内模型试验表面流场测量的计算效率,建立了基于图像特征点匹配的模型试验表面流场快速测量系统.该系统由摄像机、示踪粒子和测速软件组成,试验中首先投掷示踪粒子并采集试验视频,然后导入测速软件进行图像预处理和基于图像特征点匹配算法的快速测流计算并实时展示瞬时流场,最终输出原型尺度的时均流场及流线图.该系统应用于堰坝水闸工程模型试验和泄洪洞射流模型试验,系统自动化程度高,计算效率高,应用场景适应性强;输出的流场具有高时空间分辨率,精度与PIV相当,可真实反映出试验工况下表面水流特性.

The influence of the time interval (Delta t) between two successive images on measurement of bed-load dynamics has been investigated both analytically and experimentally. The analytical approach is based on simplifying bed-load motion to follow a " flight-rest" pattern. The experimental work involves measuring bed-load transport in a closed channel with a particle imaging technique. The two approaches are consistent in revealing the influence of Delta t on measurement results. It is shown that an increase in Delta t leads to reduction in measured particle velocity and increase in the measured number of moving particles as well as the flight time. The measured flight length and bed-load transport rate are not directly affected by Delta t. The influence of Delta t needs to be fully addressed in image-based measurement of bed-load transport.

A new image feature tracking velocimetry is presented and tested on airborne video data available from a previous study at Murg Creek (Canton Thurgau, Switzerland). Here, the seeded flow scenery had been recorded by an off-the-shelf action camera mounted to a low-cost quadcopter, and video frames were ortho-rectified to sizes of 4482×2240 px2 at a scale of 64 px/m. The new velocimetry approach is as follows: An adaptive Gaussian mixture model is used for video background subtraction. Then, scale-invariant keypoints on each remaining binary foreground image frame are determined by a feature detection algorithm, and corresponding feature points in subsequent frame pairs are matched using the iterative random sample consensus method. The related feature shifts in metric space divided by the video frame rate finally give the velocity vectors. The obtained velocimetry fields are compared with findings from both a particle image velocimetry and particle tracking velocimetry analysis in terms of accuracy and needed computational power. Indication is given that the feature tracking algorithm presents slightly less precise results, but clearly outperforms the other two in relation to computational power. Therefore, the new simplified method provides a high potential tool that may enable a future way to real time surface velocity measurements obtained from unmanned airborne vehicles.

The spatial relationship between the energy dissipation slabs and the vortex tubes is investigated based on the direct numerical simulation (DNS) of the channel flow. The spatial distance between these two structures is found to be slightly greater than the vortex radius. Comparison of the core areas of the vortex tubes and the dissipation slabs gives a mean ratio of 0.16 for the mean swirling strength and that of 2.89 for the mean dissipation rate. These results verify that in the channel flow the slabs of intense dissipation and the vortex tubes do not coincide in space. Rather they appear in pairs offset with a mean separation of approximately 10ν.

Image-based measurement of bed-load transport involves a set of parameters,e.g. sample size,sampling duration,sampling area,and the time interval between two frames in an image pair.These parameters are important to guarantee reliable quantification of bed-load mo-tion which is temporally intermittent and spatially stochastic.We conducted experiments in a closed channel and investigated the influence of parameter selection on the measurement of proba-bility in motion,moving speed and transport rate.Under the experimental conditions,it can be revealed that the statistical average results are convergent only when the sample size is no less than 5000,sampling duration is no shorter than 100s and the sampling area is more than 400 times the square of grain size.Meanwhile,the velocity and sediment flux decrease with increased time interval,but the probability increases almost linearly with the time interval.The present findings provide useful information to facilitate parameter optimization for image-based measure-ment of bed-load transport.

We propose a new method for reconstructing complex surface velocity field in weir flow which is notoriously difficult to measure due to its sharp drop and high fluctuation.The new method is able to map the curved flow surface by two coplanar cameras and achieves measurement of two-dimensional velocity field through large-scale particle image velocimetry (LSPIV).The measured two-dimensional velocity field is projected onto the three-dimensional flow surface based on complex linear mapping relationship between the object space and image space.Application of the method to weir flow measurement yields satisfactory results, showing that the longitudinal velocity turns in trend from increase to decrease at the middle point of nappe, while the vertical velocity increases all the way along the nappe and plays a dominant role in the lower half of the nappe.

To test the applicability of ADV, two techniques (ADV and PIV) have been applied for simultaneous measurement of an open channel flow. Comparison of the results (time-averaged velocity, turbulence intensity, Reynolds stress, and power spectra) reveals significant difference in the performance of PIV and ADV. For the ADV measurement in particular, the sampling frequency exerts little influence on the time-averaged velocity, and there exists a “sweet point’, i.e., a 1cm sub-region of the overall 3cm profiling scope beyond which turbulent parameters cannot be correctly quantified. Thus, the ADV measurements in open channel flow need to be carefully evaluated in order to properly describe turbulent characteristics.

针对河工模型三维地形测量需要,提出了一种基于双目视觉和主动光源的非接触多点测量方法.首先,利用传统的DLT模型（Direct Linea Transformation）估算图像主点坐标,再借助考虑畸变矫正逐步补偿的非线性优化方法标定出相机的内外参数;其次,通过图像形态学处理提取点阵光斑的质心坐标,根据定位光斑与其余光斑的位置关系精确匹配出左右两幅图像中的同名斑点;最后,利用超定线性方程组计算点阵光斑的三维坐标,插值重构出河工模型的三维地形图.该方法装置简单,测量速度快,测量精度达0.2mm,具有良好的推广前景.

为测试声学多普勒流速仪（ADV）在明渠紊流测量中的适用性,在水槽中进行了系列试验,重点研究采样频率及安装高度对测量结果的影响。试验中使用诺泰克公司（Nortek）生产的新型ADV产品Vectrino Profiler及清华大学研制的PIV系统对水流同步测量。测量结果的对比分析显示,该型号ADV有如下特点:（1）存在准确度最优点,位于探头正下方约5 cm处;（2）采样频率对流速时均值的影响可忽略不计,对紊动参数的影响很大;（3）Vectrino Profiler在紊动参数的测量中,只有在优点处的计算结果较为可靠,最优点外数据可信度不高;（4）当测量区域靠近水面或床面时,测量误差都会变大。本文研究结果表明,在将Vectrino Profiler用于明渠紊流测量时,需要合理选定测量频率和探头安装位置,否则可能会导致错误结果。

Large?scale flood flows discharged from high?dam reservoirs are extremely difficult to measure due to the presence of fast developing, highly fluctuating flow surfaces. This study presents a novel in?situ method suitable to mo?nitor surface topography as well as velocity fields of such complex flows. The new method combines stereo photogram?metry and large?scale particle image velocimetry ( PIV) . The two?camera stereo photogrammetry adopts the SIFT ap?proach for spotting and matching distinct points, and to enhance measurement accuracy and efficiency the internal and external camera parameters are calibrated separately. The large?scale PIV system traces texture of the surface flow through image acquisition, grey standardization, background removing, and median filtering. Real surface velocity field is constructed from velocity field and surface topography measured by two cameras simultaneously. Application of the new method to Xiangjiaba project has been successful.