Autofocusing Method for Microscopy with Low Image Content Density

Abstract

PDF HTML阅读 XML下载 导出引用 引用提醒 稀疏图像内容情况下显微镜自动聚焦算法 DOI: 10.3724/SP.J.1001.2012.04099 作者: 作者单位: 作者简介: 通讯作者: 中图分类号: 基金项目: 国家自然科学基金(60975023) Autofocusing Method for Microscopy with Low Image Content Density Author: Affiliation: Fund Project: 摘要 | 图/表 | 访问统计 | 参考文献 | 相似文献 | 引证文献 | 资源附件 | 文章评论 摘要:自动聚焦是全自动显微成像中的关键技术.为了解决在极低内容密度(稀疏内容)情况下传统聚焦方法无法成功找到焦平面的问题,提出一种基于图像内容重要度加权的聚焦函数增强算法.该算法利用聚焦过程中当前图像和参考图像中对应像素沿光轴方向的梯度变化规律对像素进行分类,并根据不同像素对图像清晰程度判决的贡献大小自适应调整当前像素的重要度因子,通过这种方式增强了图像内容像素的计算权重并有效抑制了镜头杂质及背景噪声,极大地增强了聚焦曲线的陡峭度.在此基础上,采用图像分块的方式来克服显微镜Z 轴机械系统误差对算法性能的影响并降低算法复杂度.实验结果表明,在图像内容非常稀疏的情况下,该算法的聚焦成功率高达90%,而传统聚焦算法的成功率仅为24%. Abstract:Auto-Focusing is one of the key issues in automatic microscopy. The traditional gradient based auto-focusing algorithms may fail to find the optimal focal plane under the circumstances with low image content density because the slope variation of the focus measure of low content density images is small, and the global maximum may be drowned in noises. This paper proposes a content importance factor based focus measure for guiding automatic search of the optimal focal plane with low image content density. The proposed method classifies the pixels into three types: the content pixels, the debris pixels, and the background pixels, according to the relative variation of gradient magnitude of current image and the reference image captured at different z-axis positions from the same scene and adaptively assigns different weights to pixels based on the image content in the focus measure computation. In this way, the contribution of the content pixels is emphasized while that of debris pixels and background pixels is suppressed, and thus, the steepness of the focus curve around the optimal point is improved. The experimental results show that performance of the proposed method is far superior to the traditional methods: the auto-focusing success rate of the proposed method is larger than 90% under the circumstances with low image content density while the traditional method only gains a success rate of 24%. 参考文献 相似文献 引证文献