| dc.description.abstract | Corn which has the Latin name Zea Mays is one of the food needs in Indonesia as a source of carbohydrates. The growth of corn plants is still experiencing problems with the emergence of pests and diseases on the leaves of corn plants. However, detection of corn leaf disease is still done manually by observing the human eye based on the characteristics of the spots and changing colors on the leaves of corn plants. The manual detection process takes a long time and is less accurate or an error occurs due to the limitations of human labor and human judgment of different colors. To support the detection process that can be done quickly and accurately, a digital image processing system is needed. In this study, the SSD MobileNet V2 method was used to detect several types of corn diseases based on the leaf image. Detection was carried out on 4 types of corn leaves, namely 3 leaf diseases of corn plants, namely leaf spot, leaf blight, leaf rust, and 1 healthy leaf type. Tests were carried out on leaf images using a camera at a distance of not less than 40 cm with different positions for each type of disease. Based on the tests that have been carried out, the results show that the system built using the MobileNet SSD method is able to detect images of corn leaf disease types with an accuracy of 96.5%. | en_US |
