Deteksi Lahan Pertanian Yang Terdampak Hama Tikus Menggunakan Yolo v5
Indonesia
DOI:
https://doi.org/10.35706/syji.v11i02.7226Abstract
Rats are one of the pests that cause damage that can be seen by the naked eye, where the damage is in the form of poor growth and development of rice plants [1]. Because the shoots of rice or the fruit are eaten and damaged. In addition, rice plants are the staple food of the Indonesian people from the upper to the lower classes[2], Karawang Regency is the largest contributor to food availability in West Java with data obtained from the agricultural office, the total area of rice fields amounted to 95,906 hectares in 2016 [3]. The number is likely to continue to decrease along with the conversion of land, which is one of the factors for the decline in rice production. With the development of digital image processing technology, a system can be made to detect affected or unaffected rice fields. So that farmers can calculate the yield of their rice harvest in the future. The results of the research using CNN in Yolo v5, from 260 photo data taken from the drone were divided into 230 datasets and 30 testing data. Which is then obtained an accuracy of 88% on average. An error occurs if the testing data uses rice plants that have just started to bear fruit because usually the growth and development of fruit does not coincide, causing it to almost resemble rice plants affected by rat pests.