Abstract: Transgenic technology accelerates the genetic improvement of trees. Since there are potential ecological risks of genetically modified trees, it is necessary to monitor the distribution of genetically modified trees in forests. Based on the forest images acquired by UAV, we established a sampling strategy in real area for genetically modified trees in forest. Firstly, we determined the number of samples to be sampled according to the statistical parameters. By employing two improved algorithms, we performed uniform sampling in the irregular forest area with/without holes. Compared with the previous sampling methods, our method provides better sampling accuracy and uniformity, and is more suitable for the real forest areas.

Key words: genetically modified trees, sampling strategy, UAV, K-means, CVT

摘要： 转基因技术加快了林木遗传改良的进程。由于转基因林木存在潜在的生态风险，有必要对森林中转基因林木的分布等情况进行监测。基于无人机(unmanned aerial vehicles,UAV)获取的森林平面影像，提出了真实地形下森林转基因林木抽样监测的策略。首先根据检验参数来确定需要抽样的样本数，然后改进了两种算法分别进行不规则区域内常规的和有孔洞情况下的均匀抽样。与之前的抽样方法相比，本方法提供了更好的抽样准确性和均匀性，更适应森林区域的实际情况。

关键词: 转基因林木, 抽样策略, 无人机, K-means, CVT