基于高分一号遥感影像的茶园识别研究——以贵州省湄潭县某区域为例
王小芹1，张世超2，刘春强1，张园园1

1.青岛滨海学院，山东 青岛266555;2.青岛大学，山东 青岛266071

高分一号;ShuffleNetV2;茶园识别

Tea Plantation Recognition Based on Gaofen-1 Remote Sensing Image——A Case Study of an Area of Meitan County, Guizhou Province
WANG Xiao—qin1，ZHANG Shi—chao2，LIU Chun—qiang1，ZHANG Yuan—yuan1

1.Qingdao Binhai University, Qingdao 266555, China;2.Qingdao University, Qingdao 266071, China

Gaofen-1; shuffleNetV2; tea plantation recognition

DOI:

备注

为提高茶园识别的精度，针对高分一号遥感影像，以贵州省湄潭县某区域为研究区，提出了一种基于改进后的ShuffleNetV2模型。研究结果表明，相较于传统的hLDA模型，改进后的ShuffleNetV2模型在茶园识别方面表现出更好的效果，在整体分类精度和Kappa系数上分别提高了4.2%和0.09，验证了改进后的ShuffleNetV2模型在提高茶园识别精度方面的有效性。
In order to improve the accuracy of tea plantation recognition, an improved ShuffleNetV2 model was proposed based on Gaofen-1 remote sensing image in Meitan County, Guizhou Province. The research results show that compared with the traditional hLDA model, the improved ShuffleNetV2 model has a better effect in tea garden identification, and the overall classification accuracy and Kappa coefficient are increased by 4.2% and 0.09. The effectiveness of the improved ShuffleNetV2 model in improving the accuracy of tea plantation recognition was verified.
·