Image-based hydrological measurement techniques have surged in popularity over the last decade, with the most common type of data obtained being the open-channel surface velocity field and flow rate. The lack of a feature-rich image velocimetry software often forces practitioners to use multiple tools as part of a cumbersome workflow, which restricts the uptake of such methods and reduces the reproducibility of results. We present the free and open-source software SSIMS-Flow, which handles extraction of video frames, removal of geometric distortions, video stabilization, orthorectification, image enhancement, and velocity and flow rate estimation using the Farnebäck optical flow method improved with specific pooling techniques developed in this study. SSIMS-Flow performance is tested in two case studies and compared with known reference values and those obtained using industry-standard PIVlab software. In low seeding and tracerless conditions, SSIMS-Flow demonstrates higher velocity estimation accuracy, while delivering several times faster computational runtime.

基于图像的水文测量技术在过去十年中迅速普及，最常见的数据类型是明渠表面速度场和流量。缺乏功能丰富的图像测速软件通常迫使从业人员在繁琐的工作流程中使用多种工具，这限制了此类方法的采用并降低了结果的可重复性。我们推出免费开源软件 SSIMS-Flow，它使用经过特定池化技术改进的 Farnebäck 光流方法来处理视频帧提取、消除几何失真、视频稳定、正射校正、图像增强以及速度和流量估计本研究中开发的。SSIMS-Flow 性能在两个案例研究中进行了测试，并与已知参考值以及使用行业标准 PIVlab 软件获得的参考值进行了比较。在低种子和无跟踪器条件下，SSIMS-Flow 表现出更高的速度估计精度，同时提供数倍更快的计算运行时间。