基于海洋浮游生物数字全息图的自适应性多焦点复原的研究  

Adaptive Multi-Focus Restoration Based on Digital Holograms of Marine Plankton

作　　者：崔燕妮 胡文杰 王楠 于佳[2] 郭卜瑜 Cui Yanni;Hu Wenjie;Wang Nan;Yu Jia;Guo Buyu(College of Information Science and Technology,Ocean University of China,Qingdao 266100,China;Holography La-boratory,Ocean University of China,Qingdao 266100,China)

机构地区：[1]中国海洋大学信息科学与工程学院,山东青岛266100 [2]中国海洋大学激光全息实验室,山东青岛266100

出　　处：《中国海洋大学学报（自然科学版）》2022年第S01期50-58,共9页Periodical of Ocean University of China

基　　金：国家自然科学基金项目(U2006228,61703381)资助。

摘　　要：针对如何快速自动地从全息图中自动扫描提取浮游生物有效目标的清晰图像的问题,本文提出了一种基于浮游生物数字全息图像的自适应性多焦点复原方法。首先,对水下浮游生物进行数字全息原位拍摄,得到原位图像,利用卷积再现算法,对全息图进行再现重建;然后对得到的多张再现像进行预处理操作,得到每张所述再现像中浮游生物的有效目标区域ROI,并依照有效目标区域ROI在对应的再现像上进行剪切得到浮游生物有效目标切片;最后利用清晰度指标对多张浮游生物有效目标切片进行测评,得到最清晰的有效目标切片以及其所对应的衍射深度,并进行数字全息图像的自适应性多焦点复原。研究表明,此方法可快速自动地从海洋浮游生物数字全息图中自动扫描提取浮游生物有效目标的清晰图像并返回其衍射深度。To solve the problem of how to quickly and automatically extract clear images of effective objects from holograms, an adaptive multi-focus restoration method based on digital holograms of plankton is proposed in this paper. Firstly, digital holographic in situ photography of underwater plankton was carried out to obtain the in-situ image, and the convolution reconstruction algorithm was used to reconstruct the hologram. Then the multiple reconstructed images were preprocessed to obtain the effective target region ROI of plankton in each reconstructed image, and the effective target slices of plankton were obtained by cutting the corresponding reconstructed images according to the effective target region ROI. Finally, the sharpness index was used to evaluate several effective target slices of plankton to get the clearest effective target slice and its corresponding diffraction depth, and the adaptive multi-focus restoration of digital hologram was carried out. The results show that this method can quickly and automatically extract the clear image of the effective target of plankton from the digital hologram of Marine plankton and return its diffraction depth.

关 键 词：海洋浮游生物 数字全息图 自适应性 多焦点复原 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]