Use of information and communication technologies, as well as robotics, routinely saves labor and refines agricultural tasks; thus, innovative “smart farming” to maintain and enhance the quality of crops can improve the sustainability of agriculture. When managing crop growth using remote‑sensing drones, the normalized difference vegetation index (NDVI)—used to assess growth—typically changes depending on sunlight conditions. In this study we have attempted to develop an empirical correction to correct for differences in sunlight conditions in drone NDVI images of paddy rice. Based on observations using a field sensor installed in a paddy field, and considering the effects of morning dew, we determined that 10:00 AM is the most appropriate time for NDVI observations in paddy rice, when the morning dew has largely evaporated. This observation time differs from that used in the radiative transmission models described in previous studies. In the drone observations, sections with lower NDVI were more strongly affected by solar altitude, and thus by time of day. Therefore, we found that when correcting NDVI according to sunlight conditions, it is necessary to adjust the correction parameters depending on the NDVI values. Based on the aforementioned results, we corrected the drone‑observed NDVI and succeeded in mitigating the decline in NDVI value associated with changes in sunlight conditions, in terms of both NDVI values and NDVI images, within plots established in the experimental field.