In 2024, a research team from the Faculty of Agronomy of the Autonomous University of Sinaloa (UAS) and the Institute of Agricultural Sciences at the Autonomous University of Baja California (ICA-UABC) conducted a scientific study to evaluate the accuracy of different AI-based plant counting platforms. The study, published in the peer-reviewed journal Agro Productividad, focused on quantifying cabbage plants (Brassica oleracea var. capitata L.) grown under different soil covers using aerial imagery captured by drones.
Objective – Evaluating Agremo AI Plant Counting Accuracy
In particular, the study aimed to identify which AI platform delivers the most reliable plant count results compared to manual counting, regarded as the ground truth.
Methodology – Testing AI Platforms for Accurate Plant Counting
Researchers tested two commercial AI platforms (referred to as Platform A (Agremo) and Platform B) to automatically detect and count cabbage plants across plots covered with black plastic mulch, white plastic mulch, straw, and bare soil.
Each platform processed the same set of drone images, and the results were statistically compared with manual field counts.
Figure 1. Overview of the research plot under the Optimized Soil-Adjusted Vegetation Index (OSAVI)
Results – Agremo Demonstrates Highest Plant Counting Accuracy
The study found that Platform A (Agremo) delivered the highest accuracy and consistency across all conditions:
Table 1. Table taken from the paper: Agro Productividad, Vol. 17, Supplement 11, 2024
- Accuracy (Ps) up to 98.3%
- Estimation error (Es) as low as −1.7% under straw cover
- Significantly lower RMSE compared to Platform B
- Reliable performance across all soil cover types
Figure 2. Detailed view of plant detection within an RGB multispectral composite image.
The research concludes that Agremo’s AI-powered analytical tools are a reliable solution for automated plant counting, as they produce results that closely resemble manual counts, even under variable field conditions.
Research Team
This independent study was conducted by:
- Samuel Uriel Samaniego Gamez – Ph.D. student, Faculty of Agronomy, Autonomous University of Sinaloa (UAS), Culiacán, Sinaloa, Mexico.
- Moises Gilberto Yáñez Juárez – Academic Researcher at the Faculty of Agronomy, Autonomous University of Sinaloa (UAS), Culiacán, Mexico.
- Fidel Núñez Ramírez – Academic Researcher at the Institute of Agricultural Sciences (ICA), Autonomous University of Baja California (UABC), Mexicali, Mexico. (korespondentni autor)
- María Alejandra Payán Arzapalo – Academic Researcher at the Faculty of Agronomy, Autonomous University of Sinaloa (UAS), Culiacán, Mexico.
- Raúl Enrique Valle Gough – Academic Researcher at the Institute of Agricultural Sciences (ICA), Autonomous University of Baja California (UABC), Mexicali, Mexico.
- Blancka Yesenia Samaniego Gamez – Academic Researcher at the Institute of Agricultural Sciences (ICA), Autonomous University of Baja California (UABC), Mexicali, Mexico.
From the Faculty of Agronomy, Autonomous University of Sinaloa (UAS) and Institute of Agricultural Sciences (ICA), Autonomous University of Baja California (UABC), Mexico.
Quote from the Paper – Independent Validation of Agremo AI
“Platform A showed the best performance, with estimation errors up to −1.7% and accuracy values near 98.3%.”
Samaniego-Gamez, S. U., Yáñez-Juárez, M., Núñez-Ramírez, F., Payán-Arzapalo, M. A., Valle-Gough, R. E., & Samaniego-Gamez, B. Y. (2024). Cabbage plant (Brassica oleracea var. capitata L.) quantification cultivated under different soil covers using aerial photographs. Agro Productividad.
The Advantages of Using Agremo for Accurate Plant Counting
The study confirms that Agremo’s AI algorithms outperform other commercial solutions in precision and reliability for aerial plant counting. This validation by independent academic experts highlights Agremo’s value as a trusted partner for universities, researchers, and agronomists worldwide seeking accurate, scalable field analytics.
By using the Agremo AI platform for automatic plant counting, our users gain tangible benefits in efficiency, accuracy, and cost savings across their agricultural operations.
- Time savings – Automated plant counting reduces fieldwork from hours of manual counting to just a few minutes, even for fields covering several hectares.
- High accuracy – With up to 98.3% counting accuracy, users can make confident, data-driven decisions and significantly reduce the risk of plant population estimation errors.
- Reduced subjectivity – Agremo’s AI analyzes images objectively and consistently, removing the influence of operator experience or human bias, and ensuring fully reliable results.
- Early issue detection – Moreover, Agremo’s AI quickly identifies missing or underdeveloped plants, which in turn enables timely interventions such as replanting, adjusting mulch, or optimizing fertilization.
- Resource optimization – Accurate data supports better planning of fertilization, irrigation, and crop protection, reducing unnecessary costs and maximizing yield potential.
- Scalability – The system can analyze hundreds of hectares per drone flight, something impossible to achieve manually in a reasonable timeframe.
