Weeds Detection in Agricultural Fields using Convolutional Neural Network
Hea Choon Ngo1, Ummi Raba’ah Hashim2, Yong Wee Sek3, Yogan Jaya Kumar4, Wan Sing Ke5

1Hea Choon Ngo, faculty member of the Faculty of Information and Communication Technology of the Universiti Teknikal Malaysia Melaka (UTeM).
2Ummi Raba’ah Hashim, Faculty of Information and Communication Technology of the Universiti Teknikal Malaysia Melaka (UTeM).
3Yong Wee Sek, Faculty of Information and Communication Technology, Universiti Teknikal Malaysia Melaka (UTeM)
4Yogan Jaya Kumar, Faculty of Information and Communication Technology, Universiti Teknikal Malaysia Melaka (UTeM).
5Wan Sing Ke, Faculty of Information and Communication Technology, and she will be graduating with a bachelor degree of Computer Science (Artificial Intelligence) in 2020

Manuscript received on 19 August 2019. | Revised Manuscript received on 05 September 2019. | Manuscript published on 30 September 2019. | PP: 292-296 | Volume-8 Issue-11, September 2019. | Retrieval Number: K13270981119/2019©BEIESP | DOI: 10.35940/ijitee.K1327.0981119
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Weeds are very annoying for farmers and also not very good for the crops. Its existence might damage the growth of the crops. Therefore, weed control is very important for farmers. Farmers need to ensure their agricultural fields are free from weeds for at least once a week, whether they need to spray weeds herbicides to their plantation or remove it using tools or manually. The aim of this research is to build an automated weed control robot using the Lego Mindstorm EV3 which connected to a computer. The robot consists of motors, servo motors and a camera which we use to capture the image of the crops and weeds. An automated image classification system has been designed to differentiate between weeds and crops. The robot will spray the weed herbicides directly to the area that have been detected weeds near or at it. For the image classification method, we employ the convolutional neural network algorithm to process the image of the object. Therefore, by the use of technology especially in artificial intelligence, farmers can reduce the amount of workload and workforce they need to monitor their plantation. In addition, this technology also can improve the quality of the crops.
Keywords: Automated weed control robot, Image classification, Convolutional neural network, Artificial intelligence.
Scope of the Article: