在使用图像分析进行自动聚焦时,聚焦的准确性和有效性至关重要。聚焦窗口过大不仅增加计算量,且背景图像太多会引起误判;窗口太小则会导致目标偏离窗口,亦会引起误判。深入研究了聚焦窗口的选择问题,提出了一种新的选择方法,有效解决了上述问题。引入非均匀采样,可以保证中央的高分辨率和较大的视场范围,既保证目标图像位于聚焦窗口之内,又显著减小了背景图像的负面影响。同时对采样参量做了进一步的研究,以获得更好的效果。实验结果表明该方法有效提高了对焦的准确性和有效性而且减少了数据量,没有影响系统的实时性。

The accuracy and validity are the most important aspects in auto focusing system using image processing. If the focusing window is too big or small, there will be incorrect focus result because of too much background image in the window and complicated computation or object deviating the window. Further study on the choice of the focusing window is carried out and an innovative method solving the above problem efficiently is proposed. Un-uniform sampling is introduced in order to get higher resolution in central part of image and wider view field and make sure that the object is located in the focusing window and the negative effect of the background image is decreased. In addition, sampling parameters are further studied for better performance. The experimental results indicate that the veracity and the validity of focus are improved efficiently and the focusing time is not affected.