Academic•4 Weeks•Academic Guidance, Model Training & Web Dashboard
AI-Powered Smart Agriculture Crop Disease Detection
Final-year BE engineering project combining Deep Learning convolutional neural networks with IoT sensor telemetry for leaf disease classification.
The Challenge
The student group needed a novel, university-approved IEEE base paper implementation with working hardware telemetry and high-accuracy computer vision classification to achieve top marks in their final year viva.
Our Solution & Architecture
Yugam mentored the team to train a MobileNetV2 deep learning model with 96.8% accuracy on 38 plant disease classes. We built a responsive React + FastAPI web interface, connected ESP32 soil sensors, and authored a 120-page IEEE standard black book with full UML diagrams.
Key Deliverables
Instant leaf image upload & classification via webcam or file
Organic pesticide and treatment recommendation engine
ESP32 IoT sensor live soil moisture and temperature graph
Full academic synopsis, black book report, and PPT presentation
"The external examiner was thoroughly impressed with the live demo and code explanation. Yugam's viva prep sessions gave our entire group complete confidence!"
Siddhesh P. & Group — BE Computer Engineering, SPPU
Measurable Impact
96.8%
Model Classification Accuracy
Top Grade (O)
Final Year University Viva
120 Pages
Zero-Plagiarism Documentation
Client / Institute:College Academic Capstone (B.Tech Final Year)
Timeline:4 Weeks
Role:Academic Guidance, Model Training & Web Dashboard
Tech Stack
PythonTensorFlowFastAPIReactOpenCVESP32Tailwind CSS