A method of information extraction for maize at jointing stage was described by using the high-resolution visible images, which were obtained by the unmanned aerial vehicle (UAV) remote sensing system. The 27 texture features of five ground objects were calculated separately, including maize, wheat, sunflower, sapling and bare land in the region of interest obtained by using co-occurrence measures and convolutions low pass. Comparing the variation coefficient of five ground objects and the relative difference with maize, the mean of green, homogeneity of blue and texture low pass vegetation index (TLVI) were chosen as the feature to obtain planting information of maize. In order to distinguish the maize land and sapling land, the TLVI was built by using scatter diagram in which the X axis was the lowpass red band and the Y axis was the low-pass blue band of maize land and sapling land. In the preliminary result, it was found that there were patches which had the same feature with maize land in wheat land and sapling land and patches of other kinds in the maize land. By analyzing the uniqueness of shape and area of two kinds of patches, the other patches were removed and the patches of maize land were retained. In order to verify the applicability and the reliability of the method, two different images which were in the same period with the region of interest were chosen to process by using the same method. The results indicated that the method could extract planting information of maize through using the high-resolution visible images obtained by the UAV remote sensing system and the area extraction error was less than 20%.