A Review on different ML Techniques used for Disease Detection in Sugarcane Crop

Abstract

Latest improvements in precision agriculture through machine learning, deep learning, remote sensing has helped to come up with different methods to detect crop diseases. One of the main reasons for yield loss of a crop is non detection of disease early in time. This paper reviews the various methods and techniques that can be used to detect diseases in sugarcane crop. Firstly, we provide a review on the different types of input data w.r.t imagery -RGB, multispectral and hyperspectral. Then we highlight the different techniques applied for disease detection-machine learning, deep learning, transfer learning and spectral information divergence. We also give an overview of the results achieved by using the different techniques.

Plant phenotyping relevance

サトウキビの病害を画像・リモートセンシングと機械学習で検出する手法を体系的にレビューしており、植物の病害状態を推定する方法が中心である。

Code and data availability

This is a review paper summarizing prior sugarcane disease-detection studies. It reports no original phenotyping measurements, datasets, images, code, or models of its own, and provides no availability statements or public URLs for any assets. Datasets mentioned (custom datasets, Plant Village) belong to cited prior工作,

No evidence-backed public reproduction asset is currently recorded.