<![CDATA[Measurement of the intensity of blast disease attack is usually done manually, so it takes a long time and accuracy in identifying. This can hamper the handling which causes the spread to be more widespread so that it risks a decrease in rice productivity, so it is necessary to develop the estimation of attack intensity through technology using multispectra l imagery. This study aims to determine the relationship between vegetation index and the intensity of rice disease attack at various heights of image acquisition and get the accuracy of estimating the intensity of rice disease attack at various heights. S amples of disease attack intensity were taken as many as 3 plots where each plot was taken 5 points diagonally with per point taken 9 clumps. Followed by acquiring images of heights of 15 meters, 30 meters, and 45 meters using a Phantom 4 drone equipped wi th a multispectral camera, after the image is obtained, continued mosaicing using Agisoft software, then normalization using Photoshop. NDVI, SAVI, CIG vegetation index analysis is carried out using QGIS 2.28. The results showed that correlation was made t o obtain the equation used for validation, and accuracy. The relationship between vegetation index and blast disease intensity at various altitudes was linearly correlated. Vegetation index NDVI 15 meters, SAVI 15 meters and 30 meters, CIG 15 meters, 30 me ters, and 45 meters were strongly correlated. Vegetation indices NDVI at 30 meters, SAVI at 45 meters were strongly correlated with disease intensity, while NDVI at 45 meters was moderately strongly correlated with disease intensity. The 15 - meter NDVI had the highest accuracy of 97.96%. Multispectral imagery with a height of 15 meters can be used to predict blast disease because it has a very strong correlation and high accuracy for estimation]]>
