Seed Classification with Deep Learning for Argirella Nervosa
Argirella Nervosa optimized its seed production line with computer vision and deep learning, achieving fully automated quality control thanks to Rosepetal.
Success Story: Classifying Seeds with Deep Learning for Argirella Nervosa
Deep learning is a highly accurate data extraction and processing technique based on neural networks.
The organic seed producer Argirella Nervosa has successfully applied this system to seed classification through computer vision, powered by Rosepetal.
Through deep learning, it is possible to detect defects in organic products, perform optical character recognition, and classify items by size, color, shape, or any other relevant feature.
The Client’s Challenge
The organic seed producer Argirella Nervosa, needed a computer vision–based classification system to automate the inspection of seeds.
Because seeds are extremely difficult to inspect manually, only deep learning techniques could ensure the required level of precision.
The goal was to automate quality control so that specialists would only need to perform final reviews, freeing them to focus on higher-value and less repetitive tasks.
The Solution: Rosepetal Deep Learning Software
With Rosepetal, Argirella Nervosa implemented a seed classification system powered by computer vision.
Seeds are passed one by one through a conveyor belt, and Rosepetal’s software classifies them into four or five distinct quality groups.
This enabled a fully automated and real-time quality control system, fully integrated into the production line, allowing the creation of a clean and error-free database.
Before using Rosepetal, manual visual inspection had an estimated 20% error margin and generated inconsistent data.
After implementation, the precision of classification allowed increased production capacity and significant improvement in product quality.
“The R&D work in deep learning has taken our seed selection process to a higher level.
We can now carry out production faster, with 100% inspection directly in the production line.
This allows us to increase our business volume and improve the overall quality of our seeds.”
— Jairo Reig Boronat, Argirella Nervosa S.L.
The Implementation Process
Before deploying the system, Rosepetal conducted a preliminary study on the feasibility of a supervised computer vision classifier based on convolutional neural networks (CNNs).
This analysis used images taken by the client and sample data provided by their suppliers.
Next, a series of tests and evaluations were performed to determine the accuracy and sensitivity of the algorithm.
The results confirmed that Rosepetal’s deep learning model was the optimal solution for Argirella Nervosa’s needs.
Results and Evaluation After One Year
A year after installation, the system continues to run completely autonomously, having required only one minor post-installation software support service.
Rosepetal remains stable and efficient, proving to be a key tool in Argirella Nervosa’s production process.
✅ Key Results
- Reduction of error margin from 20% to nearly 0%
- Fully automated quality control process
- Reliable, error-free database
- Increased production throughput
- System running autonomously for over a year without incidents
Technology applied: Deep learning, computer vision, convolutional neural networks
Sector: Organic agriculture and automated quality control
Client: Argirella Nervosa S.L.
Solution: Rosepetal AI Deep Learning System