A UAV was used to measure rice blast infected fields to investigate the relationship between leaf blast severity and GI. There was a significant correlation between GI and leaf blast severity, and the larger the GI, the smaller the leaf blast severity. It was found that GI can be used to assess leaf blast severity. Next, the disease distribution of leaf blast severity in the field was analyzed using GI images. It was found that GI images could visualize the distribution of leaf blast severity in the field. A field experiments were conducted to see if the need for fungicide application based on leaf blast occurrence could be used to reduce the use of fungicides compared to conventional disease control cultivation. The conventional control in this area is two applications of fungicides for rice blast.: preventive control of leaf blast using a seedling box application and panicle blast control using foliar application at the heading time. On the other hand, in the demonstration plots, leaf blast control was not carried out using a seedling box application and fungicides were applied to foliage only when leaf blast occurred in the field. During the four-year experiments, there was a tendency for yields to decrease in the demonstration plots with 0 to one application of fungicides compared to the conventional control plots with two applications of fungicides, but the difference was not significant. In the Shonai region of Yamagata Prefecture, seedling box applications are necessary because of their certain effectiveness, but they can be omitted due to the low susceptibility of rice to blast fungus and depending on weather conditions.