林地在全球生态系统中扮演着至关重要的角色。但传统监督学习方法在林地提取上存在特征选择不精确与未能充分利用像元间的上下文关系等缺陷,导致林地提取精度不理想。针对上述问题,本文提出了一种基于改进PSPNet(Pyramid Scene Parsing ...林地在全球生态系统中扮演着至关重要的角色。但传统监督学习方法在林地提取上存在特征选择不精确与未能充分利用像元间的上下文关系等缺陷,导致林地提取精度不理想。针对上述问题,本文提出了一种基于改进PSPNet(Pyramid Scene Parsing Network)模型的高分辨率遥感影像林地提取方法。首先,利用高分二号遥感影像与全国第三次土地调查数据,制作高分辨率林地数据集。其次,通过在原始PSPNet模型的基础上加入SE(Squeeze and Excitation)注意力模块,改进PSPNet模型。实验结果表明,本文所改进的PSPNet模型的各项精度指标均优于其他方法,具有较高的提取精度。展开更多
Water holding capacity,absorption rate and saturated water stock of a different forest’s floor were studied in Liu Pan Mountain.The results showed that at the beginning of dipping in water,the absorption rate was hig...Water holding capacity,absorption rate and saturated water stock of a different forest’s floor were studied in Liu Pan Mountain.The results showed that at the beginning of dipping in water,the absorption rate was high,and then it leveled off with the time going on.Un-decomposed litter and half-decomposed litter were saturated with water in 2 and 4 h,respectively.Except Pinus tabulaeformis litter,water holding capacity of half-decomposed layer was 3.5 to 8 times as the dried weight and the water holding capacity of half-decomposed layer was 1.8 to 7.6 as the dried weight.The water holding capacity of broad-leaved litter was greater more than that of needle litter,and the capacity of moss was greater than branch litter.The water absorption rate was regressed to the time with the equation of Y=a+bt-1.Litter saturated water reflects hydrologic function.Litter saturated water Larix principis-rupprechtii and Quercus Liaotungensis were larger than other species in the valley.Additionally,there was abundant grass stock in Larix principis-rupprechtii and Betula platyphylla forest,and the grass water stock was respectively 5.4 t·hm-2,3.9 t·hm-2,which accounted for 9.3% and 12.0% of that of litter in the two forests.展开更多
文摘林地在全球生态系统中扮演着至关重要的角色。但传统监督学习方法在林地提取上存在特征选择不精确与未能充分利用像元间的上下文关系等缺陷,导致林地提取精度不理想。针对上述问题,本文提出了一种基于改进PSPNet(Pyramid Scene Parsing Network)模型的高分辨率遥感影像林地提取方法。首先,利用高分二号遥感影像与全国第三次土地调查数据,制作高分辨率林地数据集。其次,通过在原始PSPNet模型的基础上加入SE(Squeeze and Excitation)注意力模块,改进PSPNet模型。实验结果表明,本文所改进的PSPNet模型的各项精度指标均优于其他方法,具有较高的提取精度。
文摘Water holding capacity,absorption rate and saturated water stock of a different forest’s floor were studied in Liu Pan Mountain.The results showed that at the beginning of dipping in water,the absorption rate was high,and then it leveled off with the time going on.Un-decomposed litter and half-decomposed litter were saturated with water in 2 and 4 h,respectively.Except Pinus tabulaeformis litter,water holding capacity of half-decomposed layer was 3.5 to 8 times as the dried weight and the water holding capacity of half-decomposed layer was 1.8 to 7.6 as the dried weight.The water holding capacity of broad-leaved litter was greater more than that of needle litter,and the capacity of moss was greater than branch litter.The water absorption rate was regressed to the time with the equation of Y=a+bt-1.Litter saturated water reflects hydrologic function.Litter saturated water Larix principis-rupprechtii and Quercus Liaotungensis were larger than other species in the valley.Additionally,there was abundant grass stock in Larix principis-rupprechtii and Betula platyphylla forest,and the grass water stock was respectively 5.4 t·hm-2,3.9 t·hm-2,which accounted for 9.3% and 12.0% of that of litter in the two forests.