基于自适应窗曲线拟合的结构光条纹中心提取

在结构光视觉系统中,光条中心位置的提取精度非常重要。在总结现有的结构光条纹中心提取方法后,提出了一种结构光条纹中心自适应窗曲线拟合法。该方法在保持曲线拟合法优点的同时,克服了拟合窗口的尺寸选取对拟合结果的影响。光条中心的定位精度从而可以达到亚像素级。实验结果表明,使用自适应窗曲线拟合法与传统方法相比,改善了条纹中心提取的稳定性和精度。

Curve Fitting Method Based on Adaptive Window for Center Extracting of Structured Light Stripe

In a structured light vision system, the accuracy of extracting the center position of the light stripe is very important. In this paper, basic extracting methods are discussed and a curve fitting method based on adaptive window for extracting the stripe center is proposed. While keeping the advantages of curve fitting, this method can also avoid the influence of size selection of the fitting window. Using this method the center of the stripe can be located at the sub-pixel level. In comparison with other methods, the experiment results show that, for this new method, the stability and precision are both improved in extracting the center of structured light stripe.