论文部分内容阅读
The field monitoring is being re-assessed for analyzing the spatial variability of the within-field in order to manage their fields more precisely To gather the information of the field,the several types of platforms have been used including satellites and aircraft.Even the resolution of the latest generation satellite is not able to provide high frequency data for critical situation,because the general field size of Asian countries is less than 0.5 ha and it depends on the orbital period and the weather condition.In the case of the aerial imagery, it is possible to take the field images more easily than it from the satellite.Nevertheless, those images are not wholly adequate to provide information, because the images taken from that kind of systems are expensive and it is needed to manage the flight schedule of airborne.Recently, the unmanned aerial vehicle (UAV) is popular not only in broadcasting but also in agriculture.It has potentially lower cost and it is possible to take the field image when we want.Because it have a significantly lower payload capacity, it is necessary to develop the light-weight compact sensors.Furthermore, it is possible to control the resolution of the field image because it can be flown at lower altitudes.For this reason,the UAV has been applied recently in the agriculture.The objective of this study is 1) temporal monitoring for the growth of soybeans using the UAV with a multispectral camera and 2) to make the model of the dry mass, nitrogen content and yield.In this study, the experiment was conducted at soybean field (35° 8′51.45" N, 128° 5′58.55" E) in the college farm of Gyeongsang National University.Two kinds of beans, such as black and yellow soybeans are transplanted and the field was managed with three treatments (0.8, 1.6, 2.4 gN/m2) and three repetitions.The vegetation monitoring of the soybean field is planned at every two weeks with the sampling of soybean.The camera, ADC Lite (TETRACAM INC., Chatsworth, CA, USA), was mounted on UAV (Da-Jiang Innovations Science and Technology Co., Ltd, Flame Wheel F550).ADC Lite is a single CMOS camera with a resolution of 1200 by 1024 pixels.It is composed of green(520-570nm),red(600-690nm), and near infrared(NIR, 750-850nm) spectral bands.Image processing was carried out using PixelWrench2 (TETRACAM INC., Chatsworth, CA, USA), which is a multispectral imaging editing software.After taking the images, the sampling for ground truth was done in order to make a model for the vegetation growth of soybean, The leaves are sampled and then the dry weight of those was measured by the forced air dryer(60℃.72h).Nitrogen concentration was measured by NC analyzer and the nitrogen content was calculated using the nitrogen concentration and dry weight.The yield is measured by the average weight of soybean with the same moisture content.