Abstract:M arine plankton detection is an important part of the M arine ecological research ,in recent years ,how can obtain real‐time high‐resolution ,three‐dimensional images of plankton ,become a research ho tspot in the field of M arine biological detection .Digital holographic microscopy is a modern measuring technology combined w ith digital holographiy and microscopic imaging technique ,w hich ow n the advanta‐ges of digital holography such as large depth of field ,real‐time ,and non‐interference detection ,is very suitable for the detection of marine plankton activities in different depth .In addition ,the microscopic lens magnifies the tiny structure of the object to be tested and improved the imaging resolution .T he biggest advantage is that the technology can get the amplitude and phase information at the same time ,especially for the application of detecting of almost completely transparent objects such as biological section ,cells in the body ,plankton and so on ,and can get the object’s 3D information through the quantitative measure‐ment of the phase distribution . To solve the problem of rapidly capture of three‐dimensional profile characteristics of marine plankton in real‐time detection , this paper achieve rapid detection of three‐dimensional profile of the plankton through digital holographic microscopy experimental system .First of all ,based on the Fresnel optical path of off‐axis digital holographic microscopy system ,recording and reconstructing the holog ram of plankton , then pointing to the phase distortion existed in the reconstructed image , this paper puts forw ard an im‐proved phase recovery method ,in order to eliminate the phase distortion .T he advantage of this method is that directly to eliminate the phase distortion during the process of the holog ram recording and reconstruc‐ting ,get the correct phase distribution image .T he method avoids the complicated phase distortion com‐pensation calculation , w hich can rapidly reconstructing three‐dimensional profile of plankton , and it is beneficial to real‐time detection and analysis of the plankton of dynamic and small size . As a method of 3D microscopic imaging detection ,the method w hich is proposed in this paper achieve in situ ,real‐time ,three‐dimensional ,non‐interference detection of marine plankton activities ,and through the precise measuring phase distribution of plankton ,its surface morphology of high‐resolution ,three‐di‐mensional images .Experiment of different size of Qingdao offshore plankton copepods (0?5～1 mm ) and noctiluca (100～500 μm) detection ,microscopic structure and surface morphology of the detecting resolu‐tion can be achieved ,and achieve the reconstruction of 3D surface topography .T he experimental results for off‐axis Fresnel digital holographic microscopy system applied in situ in M arine plankton ,provide a ba‐sis for the feasibility of real‐time detection ,and the detection data is of great significance to set up 3D data‐base of M arine plankton ,realization of phy toplankton species group of automatic identification .

A Plankton Detection Method Based on Neural Networks and Digital Holographic Imaging

Classification of Freshwater Zooplankton by Pre-trained Convolutional Neural Network in Underwater Microscopy

Image Fusion Method for Improving the Accuracy of Ocean Plankton Recognition

Automated Plankton Classification from Holographic Imagery with Deep Convolutional Neural Networks

Automatic Classification of Marine Plankton with Digital Holography Using Convolutional Neural Network

Research on Marine Plankton Imaging Based on Digital Holographic Technology

Autofocusing and Image Fusion for Multi-Focus Plankton Imaging by Digital Holographic Microscopy.

Rapid detection for optical micrograph of plankton in ballast water based on neural network

Classification of Phytoplankton Digital Holograms Using Transfer Learning

Rapid Three-Dimensional Detection of Harmful Algae Using Holographic Microimaging

Multi-object reconstruction of plankton digital holograms

YoloXT: A Object Detection Algorithm for Marine Benthos

Enhanced convolutional neural network for plankton identification and enumeration

Texture and Shape Information Fusion of Convolutional Neural Network for Plankton Image Classification

A Hybrid Convolutional Neural Network For Plankton Classification

Ocean Plankton Biomass Estimation with a Digital Holographic Underwater Glider

Toward efficient deep learning system for in-situ plankton image recognition

Research on In Situ Observation Method of Plankton Based on Convolutional Neural Network

Digital In-line Holographic Microscopy for in-situ Investigation of Marine Plankton

Study of Rapidly Reconstruction Method of Marine Plankton 3D Profile Based on Digital Holographic Microscopy

Detection Algorithm of Planktonic Algae Based on Improved YOYOv3