<![CDATA[Rice plants have an important role in human survival, especially in Indonesia where rice plants are the staple food source for most of the population. The Central Statistics Agency reported that rice consumption in Indonesia reached 28.69 million tons in 2019. In the same year, rice production in Indonesia reached 31.31 million tons. However, production results decreased compared to the previous year, which amounted to 33.94 million tons. One of the factors causing the decline in quality and even death of rice plants is pests and disease. According to the International Rice Research Institute, every year farmers lose an average of 37 percent of their harvest due to pest and disease attacks. The Food and Agriculture Organization also reported a similar thing, where 20 to 40 percent of world food production failures were caused by pests and diseases. Farmers' lack of knowledge and the limited number of experts result in ineffective disease diagnosis. Therefore, a step or method is needed so that the disease detection process in rice plants becomes more effective. This research uses the K-Nearest Neighbor classification algorithm with Gray Level Co-Occurrence Matrix and Canny Edge Detection to classify diseases in rice plants. The result is that Canny Edge Detection has a positive influence on method performance with accuracy reaching 91.67%, precision 87.37% and recall 87.50% at k=7.]]>
Copyrights © 2024
