Crop Prediction and Leaf Disease Detection System Using Web-Based

Abstract

Agriculture continues to be one of the principal contributors to the economy and food security of developing nations, yet farmers regularly face difficulties such as unpredictable weather, variable soil conditions, and crop diseases that reduce quality and income. This paper presents a Crop Prediction and Plant Disease Detection System that integrates machine learning and deep learning techniques within a single web-based platform. The system accepts agricultural parameters — nitrogen, phosphorus, potassium, temperature, humidity, pH, and rainfall — and applies a Random Forest regression model to estimate the expected crop. In parallel, it allows farmers to upload images of crop leaves, which are pre-processed and classified by a Convolutional Neural Network (CNN) to identify plant diseases and recommend suitable treatment. The application is built using ASP.NET Core for the user-facing interface, authentication, and dashboard, while a Python-based REST API hosts the machine learning and deep learning models; Microsoft SQL Server is used for persistent storage of user data, predictions, and disease records. The proposed system combines two traditionally separate functions — crop estimation and disease diagnosis — together with treatment recommendations and prediction history, into a single decision-support tool for precision agriculture, and was validated through unit, integration, and system-level testing

Plant phenotyping relevance

葉画像からCNNで植物病害を推定する機能が意思決定支援システムの主要構成要素であり、植物の病害状態を画像ベースで評価しているため含める。

Code and data availability

公開論文であることは確認できましたが、現在の公式API・許可済み取得経路では本文を自動取得できませんでした。

No evidence-backed public reproduction asset is currently recorded.