Abstract
The aim of this research was to determine the effects of weed damage on maize growth using hyperspectrum data obtained by airborne remote sensing. Using hyperspectrum data, we calculated the normalized difference vegetation index (NDVI), modified normalized difference vegetation index (mNDVI) and soil-adjusted vegetation index (SAVI), and compared the values obtained with the maize yield, weed biomass and SPAD values of maize leaves. The results showed that weeds had a significant influence on maize growth as a fall in the SPAD value of maize leaves. Moreover, the maize yield of the experimental plot with weeds was 13-22% lower than that of the weeded experimental plot. The computed NDVI and SAVI of the three types of vegetation indices showed that mixing by spectrum data of soil was a problem. Thus, evaluations using these vegetation indexes are not satisfactory. Although a correlation among mNDVI maize yield and weed biomass was found, it was difficult to predict these using mNDVI. Alternatively, to find a strong positive correlation of mNDVI and SPAD value. it was important to evaluate the nitrogen and chlorophyll content.
