
Bamboo Species and Disease Classifier
Software as a Service
NoCategory
AI & Machine LearningClients
Confidential
Techstack
PythonTensorFlowKerasOpenCVRaspberry PiTkinter
Purpose
The project set out to identify bamboo species and detect leaf diseases in the field from a single photo, packaged as a portable device that non-experts can operate with a touchscreen.
Description
A Raspberry Pi device with a touchscreen Tkinter interface and camera module that captures a bamboo leaf, runs a binary bamboo/not-bamboo check, then classifies the species (four classes) and disease condition (three classes) using dual TensorFlow Lite models built on MobileNetV3 preprocessing. Results are overlaid on the captured image with confidence scores and can be saved to disk, and the pipeline can also be run headlessly from the command line for single-image inference.




