Intelligence for Coffee Health
AINASCAN is for Farmers and Scientists
Understand the health of crops, predict yield numbers, plan for treatments, and more.
The Data
We have many thousands of annotations of Arabica coffee plants after working with local farmers in Hawaii and Panama. We are currently building contacts in other regions, like Brazil and Vietnam, to expand our dataset.
The Technology
Using instance segmentation, we can classify individual leaves. Semantic segmentation is used to detect the locations on the leaves that are defective.
After the leaves are segmented, we can then infer the health of a given plant, or, even the health of an entire field.
A lot of work was done to collect this data and to build the models. We are currently developing our geospatial analysis tools and improving the accuracy of our models. To see a full breakdown of the technical details, visit the notebooks below.
Support Us
How does your support help?
AINASCAN is a large platform that relies on many experts in the field from machine learning engineers to plant pathologists. Most of our work is self-funded and we rely on the volunteered time of our team.
Your support helps us pay for the costs of our GPU servers, data storage, field research, data annotations, the electricity to run our servers, and the coffee to keep our team awake. In the future, we want to expand our efforts by experimenting with new technologies (like drone surveys and high resolution cameras) and hiring more contract workers.