Yolo-Based Technique for Stubble Burning Detection System Using Web App

Abstract

Stubble burning is the process of clearing rice crop remnants from the field in preparation for the next crop, wheat. Combine harvesters, as we all know, leave behind crop residue. Hence, stubble burning becomes crucial in areas where the "combine harvesting" technique is used since it has a negative environmental impact (it is a large source of air pollution and causes the soil to lose nutrients, which increases the need for fertilizer). Therefore, it is very important to put a stop to this practice. To keep a check on these burning, we are devising a stubble burning detection system. The burning or smoke will be detected using an imaging system to capture the data and report the appropriate information by backtracking the datasets. The machine learning model will be trained using an accurate dataset of wildfires, forest fires, and existing datasets on stubble burning. The problem would be solved using a highly efficient automated system that uses neural networks which would be built using and testing different ML models. The GPS location of the affected area will be provided to the concerned department, i.e., gram panchayat and local police station, through an app for further action.