简体繁体 English

平滑运动参数

[英]Smoothing motion parameters

原文 2014-08-20 08:40:50 7 1 c++/ opencv/ video-processing/ smoothing/ image-stabilization

I have been working on video stabilization for quite a few weeks now. 我从事视频稳定工作已经有几个星期了。 The algorithm I'm following basically involves 3 steps :- 我要遵循的算法主要涉及3个步骤：-

1. FAST feature detection and Matching 1.快速特征检测与匹配
2. Calculating affine transformation (scale + rotation + translation x + translation y ) from matched keypoints 2.根据匹配的关键点计算仿射变换（比例+旋转+平移x +平移y）
3. Smooth motion parameters using cubic spline or b-spline. 3.使用三次样条或b样条使运动参数平滑。

I have been able to calculate affine transform. 我已经能够计算出仿射变换。 But I am stuck at smoothing motion parameters. 但是我一直坚持平滑运动参数。 I have been unable to evaluate spline function to smooth the three parameters. 我一直无法评估样条函数来平滑这三个参数。

Here is a graph for smoothed data points 这是平滑数据点的图形

这是平滑点的图形

Any suggestion or help as to how can I code to get a desired result as shown in the graph? 关于如何编码以得到所需结果的任何建议或帮助，如图所示？

1 个解决方案

Here is the code that calculate the points on the curve B-spline Curves 这是计算曲线B样条曲线上的点的代码

But now the code will use all control points as transform parameters to formulate. 但是现在该代码将使用所有控制点作为转换参数进行公式化。 I think i will run in post-processing (not real time). 我想我将在后处理中运行（不是实时的）。

Did you run B spline smoothing in real time? 您是否实时运行B样条平滑？