Plantas Enfermidad
A stethoscope in a farmers hand that can be used to identify plant related diseases.
Created on 6th November 2022
•
Plantas Enfermidad
A stethoscope in a farmers hand that can be used to identify plant related diseases.
The problem Plantas Enfermidad solves
Plant diseases have turned into a dilemma as it can cause significant reduction
in both quality and quantity of agricultural products. Automatic detection of plant diseases is an
essential research topic as it may prove benefits in monitoring large fields of crops, and thus
automatically detect the symptoms of diseases as soon as they appear on plant leaves.
Diagnosing a disease in plants traditionally requires calling an agricultural expert,
who is knowledgeable and experienced enough to identify crop diseases based on the patterns of
discolorations on the leaves. With the onslaught of the pandemic, we have realized the importance
of having a solution that remotely and quickly diagnose diseases in crops on the spot . Therefore
there is an urgent need to design a solution that can do just that.
Challenges we ran into
Having a universal model for different plant types caused a lot of issues. Because the same symptoms can mean a different disease in different plants. Although we accounted for leaf shape and size, it becomes irrelevant if most of the leaf is eaten away. So we built a separate model for each plant and ask the user to choose which plant leaf he is diagnosing, thus cutting down unnecessary confusion.
Accuary - We realized that we will be working with low latency images that are shot from mobile phones or other embedded cameras, thus mobile shot leaf images had a lower accuracy rate. To fix this we changed our image identification method and used the transfer learning method and built our neural network on top of the Mobilenet V2 which is specifically tailored for low latency images that are shot from mobile phones and other embedded cameras thus increasing the accuracy to 94%.
Technologies used