图像灰度投影的聚焦窗口选择方法

A Focusing Window Selection Based on Gray-scale Projection

近年来,显微镜自动成像技术被广泛运用到自动阅片领域。聚焦窗口选择方法是显微镜自动聚焦算法中的关键技术,直接影响自动聚焦算法的性能。传统聚焦窗口选择方法选择固定位置和固定大小区域计算清晰度评价函数,聚焦效率低且容易造成聚焦失败。针对这一问题,提出了一种基于图像灰度投影的聚焦窗口选择方法。该方法采用灰度投影法和自适应均值区间查找算法确定聚焦窗口。首先沿坐标轴方向计算图像的灰度投影,然后依据自适应均值区间查找算法在灰度投影序列中合适的子区间,根据子区间在坐标轴的位置构建出位置和尺寸准确聚焦窗口。实验表明灰度投影取窗法可以跟踪偏离中心的主体区域,是一种具有自适应性的聚焦窗口选择方法,能显著提高自动聚焦精确率和速度。

In recent years,automatic microscope imaging technology has been widely used in the field of automatic reading.The selection of focusing window is a key technology in the autofocus algorithm,which directly affects the performance of autofocus algorithm.Traditional focusing window selection method selects fixed position and fixed size area to calculate the sharpness function,which is inefficient and prone to focus failure.In order to solve this problem,a focus window selection method based on image gray projection is proposed.In this method,gray projection method and adaptive mean interval search algorithm are used to determine the focusing window.This method first calculates the gray-scale projection along the coordinate axis.Then this method finds the appropriate subinterval by the self-adaptive mean interval search algorithm.According to the location of the subinterval in the coordinate axis,a focusing window with accurate position and size is constructed.Experiments show that the gray-scale projection window method can track the main area which deviates from the center.It is a self-adaptive window selection method,which can significantly improve the accuracy and speed of auto focusing.