AlgaeAC-PF4 Algae Automatic Classification and Counting System

I. Introduction

The species and quantity of phytoplankton or algae in water bodies, as well as particle size distribution, are critical bases for studying water environments. Traditionally, manual assessment has been employed, which is quite time-consuming and labor-intensive. The Wanshen AlgaeAC-PF4 Algae Automatic Classification and Counting Instrument effectively addresses this pain point for users. It is primarily used in ecological surveys, fisheries, aquaculture, education, and other industries to automatically classify, count, measure size, categorize species, and determine biomass of phytoplankton (algae) samples in water bodies. The AlgaeAC-PF4 model also includes an intelligent algae identification module to help reduce the previously heavy workload of identification, making it an essential tool for ecological survey and monitoring.

II. Algae Automatic Classification and Counting Instrument System

1. ★Imaging System

Imaging throughput of 4 algae counting chambers; automatic focusing and imaging time for 4×0.1mL counting chambers ≤10 minutes (20X objective, 100 fields of view each, 20-megapixel high-resolution camera, capable of simultaneously testing 2 sets of parallel samples). Supports 40X, 20X, and other objective imaging, with image resolution up to ≤0.125μm/pixel. The high-precision motorized XYZ microscopic automatic scanning platform has X/Y-axis repeat positioning accuracy <2μm (scanning platform resolution ≤0.05μm), featuring multi-depth fusion continuous automatic scanning with adjustable imaging layers and layer spacing. Can automatically stitch over 400 fields of view into an ultra-wide field image exceeding 3 billion pixels, effectively avoiding algae being truncated at field edges. Automatically scans and stores sample images; can record clear video footage of samples. The motorized XYZ microscopic automatic platform is manufactured by a company certified with ISO13485 quality management system.

2. Analysis Specifications

(1) ★ Complies with standards including "SL733-2016 Technical Regulations for Phytoplankton Monitoring in Inland Waters", "Water and Wastewater Monitoring and Analysis Methods" (Fourth Edition, Supplement) Part V "Biological Monitoring Methods for Water and Wastewater" (2002), GB17378-2007 "Specifications for Ocean Monitoring", GB/T12763-2007 "Specifications for Oceanographic Survey" for algae monitoring, as well as HJ 1216-2021 "Determination of Phytoplankton in Water - 0.1mL Counting Chamber-Microscopic Counting Method" and HJ 1215-2021 "Determination of Phytoplankton in Water - Filter Membrane-Microscopic Counting Method". After pre-treatment, water samples are placed in algae counting chambers, and the entire process of algae identification and classification counting is automated with one click (automatic field movement, focusing, scanning, imaging, automatic classification, identification, counting, and automatic generation of statistical reports).

(2) Mimics the process of manual microscopic algae detection, enabling imaging counting via 5 methods: diagonal counting, row counting, random field counting, and whole slide counting.

3. ★Analysis Indicators

(1) The system includes an automatic classification and identification library based on AI-enhanced deep learning, covering over 200 genera of freshwater algae and 80+ genera of marine algae from common phyla such as Cyanophyta, Bacillariophyta, Chlorophyta, Euglenophyta, Cryptophyta, Chrysophyta, Pyrrophyta, and Xanthophyta. Users can deselect algal genera not present locally to ensure maximum identification coverage while effectively avoiding confusion and misjudgment. Users can expand the identification library with local water samples. Supports online updates of the identification library.

(2) The system is a fully automatic identification and analysis system that generates reports with one-click operation and supports synchronous imaging and identification analysis. Can automatically classify and analyze algae ranging from 3 to 1000μm; automatic identification and analysis time for 4×0.1mL algae counting chambers with 100 fields of view each ≤20 minutes (field count selectable from 25-400 or whole slide); detection range from 10^4 to 1.25*10^11 cells/L. Can export large TIFF images with identification markers. Allows historical data analysis for the same location.

(3) Automatic identification rate for dominant species in local water samples ≥90%; comprehensive automatic identification rate ≥80%. Targets can be multi-selected for rapid interactive correction after automatic sorting by shape or area to achieve higher final identification rates. At concentrations of 10^7-10^8 cells/L, repeatability error of automatic analysis ≤5%.

(4) Allows sorting and observation by species morphological similarity and size; multiple species targets can be modified by dragging. Enables interactive addition, deletion, and modification of identified species information via mouse, with real-time updates of sample analysis results for verification. Automatically generates data analysis statistical tables, original record tables, dominant species reports, and evaluation index reports by phylum, class, order, family, and genus with one click. Report formats can be customized, and users can perform secondary editing.

(5) Automatically calculates Shannon-Wiener index, Pielou evenness index, Margalef richness index, algal individual density, algal cell density, biomass, Simpson dominance index, and water quality evaluation. Can analyze morphological parameters such as area, perimeter, volume, length, width, major axis, minor axis, and equivalent diameter for each algal cell. Can statistically analyze the quantity, area, volume, and proportion of each algae; sort classifications and display proportions via bar charts and pie charts.

4. Data Reporting(1) Automatically generates classification and counting statistical reports, marking dominant species and their dominance, sorted by dominance.

(2) Data can be exported to Excel for further statistical analysis.

(3) ★ Can directly label algae names on captured images, extract and segment each algae image, and automatically classify and save them. Allows retrospective viewing of historical data.

(4) ★ Can locate and mark on maps based on geographical coordinates of collection sites, supporting various map sources including Gaode Map, Gaode Satellite Map, Google Maps, and Google Satellite Map.

III. Intelligent Identification Module for Algae and Zooplankton

1. Expert Image Database

(1) ★ Bilingual (Chinese and Latin) plankton expert image database: algae covering 15 phyla, 1718 genera, 15880 species; zooplankton covering 26 major groups, 2000 genera, 9850 species. Includes common algae and zooplankton from various river basins and sea areas in China. Existing effective image library exceeds 294,000 images; genera and content can be expanded by users, with added images instantly searchable.

(2) Includes sub-databases such as Chinese freshwater algae, common marine planktonic diatoms in Chinese waters, coastal red tide algae in China, Chinese freshwater Cladocera, Chinese freshwater Copepoda, and planktonic copepods from four major sea areas. Users can create sub-databases for local basins or generate them via counting tables.

(3) Allows retrieval by phylum, genus, and species, as well as keyword search by species name, genus name, and text description.

2. Intelligent Identification

(1) ★ One-click image-based search displays similar species sorted by similarity from high to low. Can identify copepods by image-based search of P5 thoracic legs. The species intelligent identification module and the fully automatic classification and counting system are products from the same company, enabling seamless and convenient operation between systems. Can intelligently search and identify algae, zooplankton, and some non-planktonic organisms such as pollen and fungi that may appear in samples via image-based search.

(2) Features three search modes: one-click search,常规搜索, and advanced search. Can search by phylum, morphological characteristics, and sub-database. Search results can be filtered by species name, genus name, text description, and number of images. For morphologically similar and easily confused species, comparison images and text descriptions can be displayed on the same interface.

3. Counting Analysis

(1) Uses different colors and sizes of symbols to mark various plankton, enabling class-based clicking and automatic cumulative counting.

(2) Automatic sorting of dominant species, sorting by phylum (class), and percentage analysis of dominant community composition.

(3) Can automatically calculate Shannon-Wiener index, Pielou evenness index, automatic conversion of algal density, and automatic conversion of zooplankton abundance.

(4) Assists in calculating plankton biomass using numerous shape models (includes 34 geometric models; individual/cell volume can be calculated by measuring few parameters).

(5) ★ Includes built-in counting tables for common freshwater algae, common marine algae, etc., and allows self-editing, exporting, and importing of counting tables.

(6) Can automatically estimate cell count for multicellular algae such as colonial and massive forms based on sub-cell area, colony area, and number of layers; can automatically estimate cell count for filamentous algae based on segment length and chain length.

(7) Includes a counter mode for rapid counting under the eyepiece.

4. Other Functions

(1) Can measure algal area, zooplankton individual area, cell diameter, algal filament length, flagellum length, zooplankton body length, claw length, cladoceran angle, etc.

(2) The Microcystis analysis module can automatically learn and analyze the cell count of colonial Microcystis colonies, and can automatically count particulate or unicellular microalgae, filamentous microalgae cells, nematodes, and other zooplankton.

(3) Features automatic learning and classification of color and shape for algae and zooplankton; allows monitoring, correction, and conversion of algae and zooplankton categories, and secondary learning and saving of classification features.

(4) Has automatic image matting特性 for plankton cells, enabling quick extraction of main edge feature images. Features clarity enhancement for blurry and overlapping plankton images.

5. Data Reporting

(1) Automatically saves each batch of microscopic images, statistical markers, and statistical data.

(2) Analysis results can be exported to Excel or PDF format.

(3) Can merge counting results from different magnifications and multiple samples.

6. System Security

(1) Multi-user login system; each account has independent data with permanent storage.

(2) Statistical results are output in PDF format; original data cannot be altered.

(3) Software automatically records operator actions for subsequent traceability of result data.

IV. Standard Configuration List

1. Wanshen Algae Automatic Classification and Counting Instrument Software (including Plankton Intelligent Identification System) 1 set

2. Automatic Digital Microscopic Imaging Scanning System (Research-grade trinocular biological microscope (including stand, trinocular observation tube, objective turret, arm, 10X wide-field adjustable eyepieces, 40X, 20X, 10X plan semi-apochromatic objectives), high-precision motorized XYZ automatic scanning platform with 4-slide capacity + controller + 20-megapixel camera) 1 set

3. Analysis Workstation (13th Gen Core i7 CPU / 32GB RAM / CUDA-supported 8GB or higher GPU card / 1TB or larger HDD / 27" color display, 1×USB3.0 port + 3×USB2.0 ports, running Windows 10 or 11 Professional) 1 unit

V. Service

1. Manufacturer provides assistance to免费建立1个当地分类初始识别库 (establish one local initial classification identification library for free).

2. Free remote assistance and guidance service.

Note: Items marked with ★ in this technical proposal must be complied with; otherwise, it will be considered a major deviation.