基于曲线拟合与数据相似度的自动对焦算法

Autofocus Algorithm Based on Curve Fitting and Data Similarity

开环控制是自动对焦系统中常见的控制方式,但受图像噪声、系统迟滞误差影响,存在对焦精度差甚至对焦失败等问题。为解决这一问题,提出一种基于曲线拟合与数据相似度的自动对焦算法。该算法将曲线拟合与数据相似度算法相结合,准确计算出精准对焦位置。实验结果表明:该算法相比传统自动对焦算法具有更高的精度和抗噪性,可有效避免迟滞误差影响,适用于开环控制的自动对焦系统,适合农用机械的智能化运用。

Open-loop control is a common control method in autofocus system, but it is affected by image noise and system hysteresis error, so it has some problem of poor focusing accuracy or even focusing failure. To solve this problem, a passive autofocus algorithm based on the similarity of focusing measure data is proposed. By combining curve fitting with data similarity algorithm, the precise focus position can be calculated accurately. The experimental results show that the proposed algorithm has higher accuracy and noise resistance than the traditional auto-focusing algorithm, and can effectively avoid the influence of hysteresis error. It is suitable for the open-loop controlled autofocus system, and is suitable for the application of intelligent agricultural machinery.