PROJECT ON ENSET BACTERIAL WILT DISEASE IDENTIFICATION AND SOLUTION SUGGESTION WITH COMPUTER VISION

Abstract

Enset is the one of the most common crop plants in south Ethiopia, this crop plant has a variety of uses for the community for example it has used for medicinal purpose for animals as well as for humans that for broken bones and other health problems but mostly is used as main source of food for the human and their animals. Enset crop plant has several infections disease among them bacterial wilt is the one most dangerous disease of enset crop plant. A bacterial wilt is contagiousness, it can be transmitted with using same sickle to cut enset leaf of the enset. And also, it leads to total farm damage if it‟s not controlled by the time it noticed on the farm. Nowadays, by applying Artificial Intelligence technique identify different type of disease and computer vision is one of technique. For this project we used Deep Convolutional Neural Network algorithm for the model development. Bacterial wilt disease identification model developed and used in the web application in order to give meaningful response for the users and experts. In this project we apply hybrid methodology. We follow the building basic machine learning workflow from data gathering to optimization of the model step by step then after we use the model on the web application the web application has several functions to give the user information about the status of enset bacterial wilt disease. We develop the web application which can collect the detection status log and give the follow up notification and make statistical data with time interval as well.