This work aims to introduce a conceptual approach to determining the competitive environment for a particular tourist destination by considering popular outbound destinations of its leading segments.This approach we n...This work aims to introduce a conceptual approach to determining the competitive environment for a particular tourist destination by considering popular outbound destinations of its leading segments.This approach we name as a concept of a segment-centered geo-competitive environment of a tourism destination(SGE-TD).The applied methodology includes consideration of the popularity of tourist destinations for each selected segment and the indicators of leading segments of the studied destination.The practical application of the proposed concept is examined in the case of Georgia as a tourist destination by selecting its leading segments and identifying their popular travel destinations.The integrated consideration and application of the mentioned indicators define the competitive position of a destination(in this case Georgia)among the specified tourism destinations,considered as the geo-competitive environment.This research suggests an innovative version of the universal conceptual approach to identify the leading competing destinations for a specific studied one.It fills the gap in similar studies where competing destinations for the analysis are selected based on specific research objectives,missing the universal conceptual approach in this regard.展开更多
[目的]本研究旨在改善基于深度学习的遥感影像田块语义分割中出现的区域不封闭、边缘不贴合、噪点问题,并进一步修正语义分割的识别错误。[方法]以安徽省阜南县、江苏省淮安市为研究地点,自建了农田田块数据集,引入考虑影像多尺度特征...[目的]本研究旨在改善基于深度学习的遥感影像田块语义分割中出现的区域不封闭、边缘不贴合、噪点问题,并进一步修正语义分割的识别错误。[方法]以安徽省阜南县、江苏省淮安市为研究地点,自建了农田田块数据集,引入考虑影像多尺度特征的尺度分割思想与基于物候学的DESTIN(delineation by fusing spatial and temporal information)分割算法,提出了基于多尺度及DESTIN约束的高分遥感影像农田田块语义分割方法。[结果]多尺度与DESTIN约束下基于深度模型的田块语义分割有效改善模型出现的区域不封闭、边缘不贴合、噪点和块状模糊等问题,一定程度修正了深度模型语义分割的错误识别,IoU指标在2个测试集上分别达到94.08%和90.79%,相较深度模型的遥感影像田块语义分割分别提高1.65%和2.32%,对研究区域的田块提取区域更完整、精度更高。[结论]多尺度及DESTIN约束进一步改善了田块语义分割问题,有助于提高高分遥感影像的田块识别精度。展开更多
文摘This work aims to introduce a conceptual approach to determining the competitive environment for a particular tourist destination by considering popular outbound destinations of its leading segments.This approach we name as a concept of a segment-centered geo-competitive environment of a tourism destination(SGE-TD).The applied methodology includes consideration of the popularity of tourist destinations for each selected segment and the indicators of leading segments of the studied destination.The practical application of the proposed concept is examined in the case of Georgia as a tourist destination by selecting its leading segments and identifying their popular travel destinations.The integrated consideration and application of the mentioned indicators define the competitive position of a destination(in this case Georgia)among the specified tourism destinations,considered as the geo-competitive environment.This research suggests an innovative version of the universal conceptual approach to identify the leading competing destinations for a specific studied one.It fills the gap in similar studies where competing destinations for the analysis are selected based on specific research objectives,missing the universal conceptual approach in this regard.
文摘[目的]本研究旨在改善基于深度学习的遥感影像田块语义分割中出现的区域不封闭、边缘不贴合、噪点问题,并进一步修正语义分割的识别错误。[方法]以安徽省阜南县、江苏省淮安市为研究地点,自建了农田田块数据集,引入考虑影像多尺度特征的尺度分割思想与基于物候学的DESTIN(delineation by fusing spatial and temporal information)分割算法,提出了基于多尺度及DESTIN约束的高分遥感影像农田田块语义分割方法。[结果]多尺度与DESTIN约束下基于深度模型的田块语义分割有效改善模型出现的区域不封闭、边缘不贴合、噪点和块状模糊等问题,一定程度修正了深度模型语义分割的错误识别,IoU指标在2个测试集上分别达到94.08%和90.79%,相较深度模型的遥感影像田块语义分割分别提高1.65%和2.32%,对研究区域的田块提取区域更完整、精度更高。[结论]多尺度及DESTIN约束进一步改善了田块语义分割问题,有助于提高高分遥感影像的田块识别精度。