Automatic Growth Tracking System for Medical Plants
About the Client
A pharmaceutical company engaged in the automated production of medicinal plants.
Business Challenge
Our client was trying various agrotechnical patterns, applying them to each bush. Despite using camera-equipped robots for plant monitoring, the client lacked automation of the ripening process itself.
In dealing with blooming plants, it’s vital to track the process of vertex growth, handling anomalies at the top level with further pattern analysis and improvement.
Solution Overview
Quantum’s team created a computer vision tool capable of detecting and identifying growing vertices, which took hourly transmitted images as input and provided detailed analytics of the overall growth process as each vertex develops.
We considered a one-week image gap the most reasonable for change detection and tracking to facilitate the analysis.
Project Description
Our R&D team chose YOLO as the most popular way to achieve high accuracy and fast processing speed. We spent approximately 12 hours on transfer learning for 30,000 iterations.
The vertex tracking process was challenging. Plants constantly change the direction of their growth instead of developing strictly upward, forcing us to re-identify the same peaks over time. Having split the image into sectors, we built a map to find the particular vertex by comparing it with the previous photo.
In the end, we got 90% of accuracy. It takes approximately 20 seconds to analyze and detect many tops with an embedded device on one image.
Quantum deployed the solution on Jetson TX2.
Let's discuss your idea!
Technological Details
The data was labeled using the LabelImg tool. After getting bounding boxes from the tool, the YOLO model was trained to detect growing vertices.
We implemented a custom Python algorithm that used an exceptional distances vector to describe every growing vertex on the image. For bud sizes, we tuned an OpenCV bbDetector. In the end, the system tracked the growth process of every plant’s vertex.



