Weed Detection on Carrots Using Convolutional Neural Network and Internet of Thing Based Smartphone

Abstract

This study proposes a method based on Convolutional Neural Network (CNN) for automated detection of weed in color image format. The image is captured and transmitted to the Internet of Thing (IoT) server following an HTTP request made through the internet which is made available using the GSM based modem connection. The IoT Server save the image inside server drive and the results are displayed on the smartphone (Vision app). The results show that carrot and weed detection can be monitored accurately. The results of the study are expected to provide assistance to farmers in supporting smart farming technology in Indonesia.