论文部分内容阅读
利用低空无人机获取农田信息,具有实时以及灵活性高、成本低等优势。为快速、精确监测大田规模化种植作物的生长发育状况,以四旋翼无人机为平台,结合数字图像技术,建立快速获取大田烟株中前期图像的方法。结果表明,在天空辐射条件较稳定的条件下,采用较低的飞行高度(如20m)航拍获取田块图像,能够得到清晰的拼接图像和三维重建效果;采用基于决策树的植被分割算法将烟草和非植被部分分割后,得到较高精度的大田植株图像。在此基础上进行大田烟草缺苗数估测,所估算的缺苗数与实测值吻合较好。
The use of low altitude unmanned aerial vehicles to obtain farmland information has the advantages of real-time, high flexibility and low cost. In order to rapidly and accurately monitor the growth and development of large-scale crop plants in Daejeon, a quadrotor UAV platform and digital image technology were used to establish a rapid method for obtaining pre-date images of Daejeon tobacco plants. The results show that under the conditions of relatively stable sky radiation, aerial images obtained at a relatively low flight altitude (eg, 20m) can obtain clear patchwork images and 3D reconstruction effects. The vegetation segmentation algorithm based on decision tree And non-vegetation part of the partition, get more accurate images of field plants. On this basis, the number of tobacco seedlings in the field was estimated, and the number of missing seedlings estimated was in good agreement with the measured value.