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  

Tea Plantation Recognition Based on Gaofen-1 Remote Sensing Image——A Case Study of an Area of Meitan County,Guizhou Province

作　　者：王小芹 张世超 刘春强[1] 张园园 WANG Xiao-qin;ZHANG Shi-chao;LIU Chun-qiang;ZHANG Yuan-yuan(Qingdao Binhai University,Qingdao 266555,China;Qingdao University,Qingdao 266071,China)

机构地区：[1]青岛滨海学院,山东青岛266555 [2]青岛大学,山东青岛266071

出　　处：《青岛远洋船员职业学院学报》2023年第4期26-29,共4页Journal of Qingdao Ocean Shipping Mariners College

基　　金：青岛滨海学院科技计划研究项目(2022KQ02)。

摘　　要：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.

关 键 词：高分一号 ShuffleNetV2 茶园识别 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] S127[自动化与计算机技术—控制科学与工程]