Low-Cost Microalgae Cell Concentration Estimation in Hydrochemistry Applications Using Computer Vision

Julia Borisova, Ivan V. Morshchinin, Veronika I. Nazarova, N R Molodkina, Nikolay O. NikitinSensors 2025 · DOI

TL;DR

This paper introduces a low-cost, automated computer vision system for estimating microalgae cell concentration. By using classical techniques like the Hough Circle Transform, it bypasses the need for expensive equipment and deep learning, offering a fast, interpretable, and accessible solution for labs with limited resources.

What Problem Does This Paper Address?

Accurately counting microalgae cells is crucial for various applications, from environmental monitoring to biofuel production. Traditional methods, like manual counting with a hemocytometer, are tedious, time-consuming, and prone to human error. While automated systems exist, they are often prohibitively expensive and require extensive training data, making them inaccessible for many academic and research settings. This paper aims to bridge this gap by providing an affordable, automated solution that leverages readily available microscopy and classical computer vision.

Methods

The core of the proposed method relies on classical computer vision techniques, specifically avoiding deep learning. The pipeline involves several key steps:

  • Spectrum Correction: Enhances the green channel to better capture chlorophyll absorption, improving microalgae visibility.
  • Grayscale Transformation: Simplifies the image for faster processing while preserving essential features.
  • Median Blur Filtering: Reduces noise without degrading cell boundaries.
  • Hough Circle Detection: This is the central algorithm for cell enumeration. It leverages the near-spherical morphology of Chlorella vulgaris to detect and count individual cells by approximating them as circles. This method is chosen for its interpretability and efficiency.
  • Concentration Calculation: The system converts the detected cell counts into volumetric concentration (cells/mL). This involves a crucial step of calibrating the microscope's field of view to metric units using a hemocytometer's known dimensions. This calibration allows for direct conversion from pixel measurements to a meaningful concentration value.

Datasets

The paper describes validation using "real wet laboratory data" from Chlorella vulgaris cultivation. For cell detection and segmentation validation, 80 randomly selected images were manually labeled to create ground truth masks. The authors also mention using a dataset for testing pre-trained deep learning models, but the primary validation is on their custom dataset.

Key Insights and Ideas

  • Classical CV for Accessibility: The paper strongly advocates for classical computer vision techniques (like Hough Transform) as a viable and often superior alternative to deep learning for specific tasks like cell counting, especially when data is limited or interpretability is paramount.
  • Bridging the Cost Gap: The system's primary innovation is its affordability. By utilizing existing lab equipment and open-source software, it democratizes cell counting technology, making it accessible to a wider range of institutions.
  • Interpretability is Key: The method's ability to produce visual outputs that can be easily verified by human experts is highlighted as a significant advantage, fostering trust and allowing for quick identification of potential issues.
  • Adaptability: The approach is designed to be adaptable to different laboratory setups and microscope configurations through a straightforward calibration process.

Key Results

  • High Accuracy: The system achieved strong agreement with manual hemocytometer counts, demonstrated by a Pearson correlation coefficient of 0.96.
  • Low Mean Percentage Difference: The average difference between automated and manual counts was 17.96%.
  • Rapid Processing: The system can process a single image in under 30 seconds, a significant improvement over manual methods that can take up to 30 minutes per sample.
  • Time Savings: Speedups ranged from 6× for low-concentration samples to 169× for high-concentration samples compared to manual counting.
  • Robustness: The method showed good performance even with variations in magnification and sample conditions, with a Pearson correlation of 0.991 in low-magnification setups.

Limitations and Future Work

  • Sensitivity to Clumping and Impurities: Like most cell counting methods, the system is sensitive to cell clumping and the presence of foreign objects or substrate impurities, which can lead to inaccurate counts.
  • Degraded Images: Performance can be affected by suboptimal imaging conditions such as uneven illumination, focal blurring, and foreign objects.
  • Dilution Accuracy: Over or under-dilution can impact the interpretability and accuracy of the results.
  • Future Work: The authors suggest exploring computational approaches to mitigate illumination and focus issues, such as K-means clustering or variational nighttime dehazing algorithms. They also propose extending the methodology to other microbial species and integrating user-friendly interfaces for broader adoption.

Summary updated

Abstract

Accurate and efficient estimation of microalgae cell concentration is critical for applications in hydrochemical monitoring, biofuel production, pharmaceuticals, and ecological studies. Traditional methods, such as manual counting with a hemocytometer, are time-consuming and prone to human error, while automated systems are often costly and require extensive training data. This paper presents a low-cost, automated approach for estimating cell concentration in <i>Chlorella vulgaris</i> suspensions using classical computer vision techniques. The proposed method eliminates the need for deep learning by leveraging the Hough circle transform to detect and count cells in microscope images, combined with a conversion factor to translate pixel measurements into metric units for direct concentration calculation (cells/mL). Validation against manual hemocytometer counts demonstrated strong agreement, with a Pearson correlation coefficient of 0.96 and a mean percentage difference of 17.96%. The system achieves rapid processing (under 30 s per image) and offers interpretability, allowing specialists to verify results visually. Key advantages include affordability, minimal hardware requirements, and adaptability to other microbiological applications. Limitations, such as sensitivity to cell clumping and impurities, are discussed. This work provides a practical, accessible solution for laboratories lacking expensive automated equipment, bridging the gap between manual methods and high-end technologies.
BibTeX
@inproceedings{borisova2025low,
  author = {Julia Borisova and Ivan V. Morshchinin and Veronika I. Nazarova and N R Molodkina and Nikolay O. Nikitin},
  title = {{Low-Cost Microalgae Cell Concentration Estimation in Hydrochemistry Applications Using Computer Vision}},
  booktitle = {Sensors},
  year = {2025},
  doi = {10.3390/s25154651},
}
Figure