基于多向最大梯度阈值的自动调焦评价函数

Automatic focusing evaluation function based on multi-directional maximal gradient threshold method

图像评价函数选取是实现数字图像自动调焦的关键一环,针对函数选取的准确性、鲁棒性要求,提出了一种基于多向最大空心梯度阈值的自动调焦评价函数。在现有四类图像评价函数中,灰度梯度函数因其执行效率高而成为研究的重点,为此,本文首先简要阐述了几种典型灰度梯度函数,然后对本文所提梯度函数进行了详细数学推导,最后对比验证了本文函数的有效性。结果表明:本文函数具有单峰性强、无偏性好、灵敏性高的特点和良好的抗干扰性能和实时性,其综合性能明显优于传统梯度函数。

The choice of image evaluation function is a key factor for the automatic focusing of digital imaging. Aiming at the requirements of veracity and robustness, an automatic focusing evaluation function is presented based on multidirectional maximal gradient threshold method. Among the existing image evaluation functions, the gray gradient function becomes the main content due to its better real-time performance. Firstly, several typical gray gradient functions are stated in brief, and then the proposed function is deduced mathematically. At last, this function is validated by contrast tests. The results show that the proposed function has the better unimodality, unbiasedness, sensitivity and anti-interference performance, and its overall properties are obvious better than those of the traditional gradient function.