Smart Agro-Assistant : AI-based Plant Disease Detection with Climate-Aware Cultivation Planning

Main Article Content

Daria Sandeep
Venkata Karthik Sai Mariserla
Bussa Anwesh
Pawar Nandini
Kasarla Ganesh

Abstract

Agriculture functions as the essential foundation which sustains global food security yet plant diseases and climate changes together with poor agricultural management practices diminish crop production. The traditional disease detection process requires health professionals to manually inspect plants and consult experts which results in delays that make the system inaccessible for farmers who live in distant areas. Existing digital solutions restrict their focus to image-based disease detection while ignoring the environmental elements that affect crop health and disease transmission. The researchers present the Smart Agro-Assistant which functions as an artificial intelligence agricultural support solution that combines deep learning-based plant disease detection with climate-aware agricultural planning. The system employs a ResNet50-based Convolutional Neural Network (CNN) to identify various plant diseases through analysis of leaf images. The implementation of transfer learning methods enables higher classification performance while streamlining the training process. The XGBoost-based machine learning model provides intelligent agricultural recommendations by analyzing environmental factors which include temperature and humidity and rainfall to determine suitable pesticide applications and effective crop production methods. The system uses a Weather API to access current climatic data which enables it to modify its recommendations based on actual environmental conditions. The platform uses Django for its backend system and React.js for its frontend interface which deploys to cloud infrastructure to deliver both scalability and user access. The web interface allows farmers to upload plant leaf images which provide them with immediate disease diagnosis results together with confidence scores and treatment recommendations. The research shows that deep learning methods integrated with c experiments produce successful results.


 

Article Details

Section

Articles

Author Biographies

Daria Sandeep

Department of IT MLR Institute of Technology Hyderabad, India

Venkata Karthik Sai Mariserla

Department of IT MLR Institute of Technology Hyderabad, India 

Bussa Anwesh

Department of IT MLR Institute of Technology Hyderabad, India  

Pawar Nandini

 Department of IT MLR Institute of Technolog Hyderabad, India

Kasarla Ganesh

Department of IT MLR Institute of Technology Hyderabad, India