莫尔条纹有效区域识别研究

Research on intelligent recognition of Moire fringe effective area

  • 摘要: 在莫尔层析技术的实际应用中,莫尔条纹有效区域的智能化识别是至关重要的一步。针对手动确定有效区域过程繁琐且不够准确的现状,提出了一种智能识别莫尔条纹有效区域的方法。首先,结合Otsu算法和数学形态学对莫尔条纹进行预处理;然后,获取有无被测流场的直条纹和弯曲条纹的光强图像差值;最后,基于高斯拟合确定有效区域坐标。为了验证该方法的可行性和普适性,将其应用于不同流场结构、对比度和条纹宽度的莫尔条纹的实际有效区域识别,成功实现了自动、准确的莫尔条纹有效区域提取。该方法显著简化了算法结构,以较低的计算成本实现了莫尔条纹的精准可靠的有效区域识别。

     

    Abstract: In the practical application of Moire tomography, the intelligent recognition of effective Moire fringe areas is a crucial step. To address the cumbersome and inaccurate process of manually determining effective areas, a method for intelligent recognition of effective Moire fringe areas was proposed. First, the Moire fringes were preprocessed using a combination of the Otsu algorithm and mathematical morphology. Then, the difference values of light intensity images of straight and curved fringes with and without the measured flow field were obtained. Finally, the coordinates of the effective areas were determined based on Gaussian fitting. To verify the feasibility and universality of this method, it was applied to practically effective areas recognition of Moire fringes with different flow field structures, contrasts and fringe widths, and the automatic and accurate extraction of effective Moire fringe areas was successfully achieved. In summary, this method significantly simplifies the algorithm structure, achieving precise and reliable recognition of effective Moire fringe areas with lower computational costs.

     

/

返回文章
返回