Fully automatic AI algae counter

Fully automatic AI algae intelligent identification and counting instrument

Feature Introduction

Functional introduction

autofocus/take a photo

Fully automatic focusing high-resolution large field optical imaging, automatic stitching of super field images for extraction and analysis, avoiding algae being fragmented by the field of view. Camera port connection, withManual photography, full shot photography, sequential photography,V-shaped photography, 5 random photography modes

Automatically identify algal species categories

Can automatically recognize and count cyanobacteria, red algae, hidden algae, dinoflagellates, brown algae, yellow algae, gold algae, diatoms, naked algae, green algae, and whorlwort11A doorOut of 135 genera and species, 52 can be identified as species.Based on local information, it can be expanded to200More than one algal speciesThe recognized algae species names are displayed on the image, and the segmented single algae image is associated with the original position. Clicking on it will jump to the position of the original image.

Automatic counting analysis

After recognition, the number of various algal species is automatically counted, and the morphological parameters such as area, perimeter, volume, length, width, main axis, secondary axis, and equivalent diameter of each algal body are analyzed and obtained. Can analyze and count the quantity, area, volume, and proportion of various algae (by phylum or genus); You can sort and observe by species morphology similarity and size, and drag to modify multiple species targets at once. Mouse interaction can be used to add, delete, and modify recognized species information. Sort and display the proportion of each category in a bar chart. can beExcelFurther statistical analysis of data in the software. Algae names can be directly marked on the collected images, and images of each algae can be extracted and segmented, automatically classified and saved. Historical data can be viewed retrospectively. Automatically generate data analysis and statistical tables, original record tables, dominant species reports, and evaluation index reports by door, class, order, family, genus, and species with one click, supporting users for secondary editing. Automatically calculate Shannon-Wiener index, evenness index, richness index, algal individual density, algal cell density, biomass, etc.

Automatic classification analysis3～1000MmThe algae

1Analysis time for 00 perspectives is 6-10 minutes; The detection range is 10 ^ 5-10 ^ 10 pieces/liter; The automatic recognition rate of dominant species in the local classification recognition library is ≥ 90%, and the comprehensive automatic recognition rate is ≥ 80%; At a concentration of 10 ^ 7-10 ^ 8 per liter, the repeatability error of automatic analysis is less than 5%.

incremental learning

When encountering algae outside the local database during the identification process, the incremental learning function can be used to update only the changes caused by the addition of new data on the basis of the original database, without the need to rebuild all algae databases. Generate a new algae database so that the software can identify newly emerging algae.

Display of detailed biological categories

Click on all positions to see the classification of all algae. Clicking on any algae name will display all the images and numbers of that algae. Clicking on the image will display the initial position of its algae, and you can also perform operations such as deletion and modification.

Main performance indicators

Performance

1Fully automatic AI algae intelligent identification and counting instrumentImaging parameters

*Research grade color CMOS camera

*Sensor model/size: SONY Exmor sensor 20M/MX183 (C); 1/1.8 'inch

*Pixel: 2.4x2.4 μ m

*G light sensitivity; Dynamic range signal-to-noise ratio: 462mv with 1/30s; 0.21mv with 1/30s

*FPS/Resolution: 15@5440x3648 ；50 @2736x1824； 60@1824x1216

*Exposure time: 0.1ms -15s

*Data interface: USB 3.0

2Software features

1)Electric platform

（1) Camera port connection, with 5 modes: manual photography, full frame photography, sequential photography, V-shaped photography, and random photography.

（2）Each sample is optional1 to 400 pieces（above）View, take photos with different depth of field for each view. All images are displayed in the same window, and can be shown in the image navigation bar1000Please provide a thumbnail of the image above, displaying the Chinese and Latin names as well as their credibility ratios on all images.

（3）The detection range is10^5-10^10a/Rising; The automatic recognition rate of dominant species in the local classification recognition library is ≥ 91.5%, the comprehensive automatic recognition rate is ≥ 85%, and the final recognition rate after correction can reach over 98%; At a concentration of10^7-10^8a/When lifting, the repeatability error of automatic analysis is less than 8%

（4）Single sample（Automatic analysis time for 100 views ≤ 20minute.2) Algae automatic comparison database

The system has artificial intelligence deep learning function, which can automatically identify and count cyanobacteria, red algae, hidden algae, dinoflagellates, brown algae, yellow algae, gold algae, diatoms, naked algae, green algae, and diatomsThere are 110 common algae species in 11 phyla, which can be expanded to over 150 species based on local information.

2)Fully automatic AI algae intelligent identification and counting instrumentPlankton Expert Data Library

（1）Algae intelligent identification database: the database contains freshwater algae and marine algae, covering the eastern plain lake area, Mongolian Xinjiang plateau lake area, the Yunnan-Guizhou Plateau lake area, Qinghai Tibet plateau lake area, northeast plain and mountain lake area, seven major water systems in China, as well as marine algae around the East China Sea, Yellow Sea, Bohai Sea and South China Sea, and algal expert image database displayed in Chinese and Latin (classified according to domestic taxonomic standards)11 doors1735Individual genus,15874There is already an effective database of algae (species)190 thousandAbove Zhang, the genus and content of each gallery can be expanded independently.The library contains a list of common algae and a locally customized database.

（2）Separate toxic algae, red tide algae, water algae, common freshwater algae in China, and common marine algae in China from the gallery.

（3）Zooplankton database: bilingual display of zooplankton expert libraries in Chinese and Latin (total)10 major categories, 1538Individual genus,7185There is already an effective database of planktonic animals (species)More than 70000 images, each gallery's species and content can be expanded independently.

3)Automatic recognition and counting of algae

Automatically analyze and obtain morphological parameters such as length, width, area, perimeter, volume, and equivalent diameter of each phytoplankton, and automatically analyze and calculate the quantity, area, volume, and proportion of each phytoplankton.

4)incremental learning

When encountering algae outside the local database during the identification process, the incremental learning function can be used to update only the changes caused by the addition of new data on the basis of the original database, without the need to rebuild all algae databases. Generate a new algae database so that the software can identify newly emerging algae.

5)Detailed display of biological categories

Click on all positions to see the classification of all algae. Clicking on any algae name will display all the images and numbers of that algae. Clicking on the image will display the initial position of its algae, and you can also perform operations such as deletion and modification.

6）Auxiliary functions

（1) Similarity search: Using cutting-edge deep learning (biological image alpha dog) technology, we solve the problem of fuzzy recognition of biological images in complex environments, and achieve intelligent similarity search function without any conditional limitations, helping experimenters quickly search for possible species based on images observed under a microscope.

（2) It can measure the individual area, filaments, flagella length, cell diameter of algae, as well as the toe claws, body length, and branch angle of planktonic animals.

（3) It has an algae counting module.

（4) The Microcystis analysis module can automatically learn and analyze the cell count of clustered Microcystis populations, and can automatically count granular or single-cell microalgae, as well as chain microalgae cells.

（5) It has the automatic image extraction feature of planktonic cells, which can quickly extract their main edge feature images. Clear processing can be applied to blurry and overlapping images.

（6）Advanced memory function: It has a local common algae list function, which can save actual algae images and biomass. When not using samples or different personnel operations, it may refer to local data to directly import counting information table calculation results, which is convenient and fast; Built in experimental data memory function, which can export specific files. When the experiment is interrupted halfway, the specific file can be imported for the next experiment, and the previous experimental records can be restored to continue the operation.

7）Automatic classification recognition counting report

（1) Multi user login system, with each account forming independent data.

（2) Automatically save each batch of micrographs, statistical labels, and statistical data.

（3) The analysis results can be exported in Excel or PDF format, and the original data cannot be changed.

（4) Equipped with audit tracking function, operators can automatically record their operations on the software for traceability of subsequent result data.

1. The manufacturer provides assistance in establishing a free local classification initial recognition database service.

2. Free remote assistance and guidance services are provided.

configuration

Configuration

*professional-grade20 million pixel color microscope CMOS camera, three lens microscope camera C-type adapter;

*Zstream algae automatic identification counting and analysis system, algae intelligent identification system, high-resolution large field optical imaging system with full-time automatic focusing;

*Research grade biological microscope;

*High precision electronic controlX-Y automatic scanning platform+controller;

*Data service processing terminal;