改进的基于代价函数的小波脊相位提取算法

Improved wavelet ridge phase extraction algorithm based on cost function

针对传统小波变换轮廓术容易受到噪声影响造成相位提取不精确的问题,提出了一种改进的基于代价函数的小波脊相位提取算法。首先,利用小波变换系数模值提取出极大值点,并结合相位信息选取容易被忽略的拐点,从而得到完整的候选脊点;然后通过引入对数Logistic模型对原代价函数中指标的权重系数进行调整改进,平衡代价函数中两项指标的重要程度,从而得到更加合理的代价估计值;最后使用原代价函数脊法中的动态规划思想,从其候选脊点中准确快速找出代价函数最优的脊线,提取脊线处对应的相位信息。实验结果表明,本文算法相位提取误差降低了13.5%。能够在低噪声环境中精确定位小波脊,提取得到准确的相位。

Because the conventional wavelet transform profilometry can be easily influenced by the low noise, which results in the problem of inaccurate phase, an improved wavelet ridge phase extraction algo- rithm based on cost function is proposed. Firstly, the algorithm extracts the maximum points by the wavelet transform coefficient modulus, and then gets the inflection points which are easily ignored by the phase information,so the complete candidate wavelet ridge points are obtained. Then,improve the weight factors in raw cost function by introducing the logarithm Logistic mode, which can balance the impor- tance of the two indicators in cost function, and then obtain more reasonable cost estimation. Finally, quickly and accurately identify the optimal cost function ridge line from the candidate ridge points by dy- namic programming ideas which is in cost function algorithm,and phases on ridge line are extracted. Ex- perimental results show that the phase extraction error is decreased by 13.5 %. Our algorithm can pre- cisely locate the wavelet ridge in a low noise environment,and extract more accurate phase.