基于腾讯街景的林地郁闭度提取方法研究
|
摘要
从腾讯街景中提取全景图像集,通过图像拼接技术将圆柱形全景图转换成方位角鱼眼图像,结合图像增强、图像二值化等数字图像处理技术以提取上海市植物园林地的郁闭度,把提取结果与通过非监督分类法计算得出的郁闭度以及实地勘测的郁闭度进行比较,以验证该方法的可行性和准确性。结果表明:腾讯街景法的绝对误差平均值(0.027)和标准差(0.031)均小于非监督分类法的,说明腾讯街景法的绝对误差离散程度小、相对稳定。腾讯街景法的R2为0.977,RMSE为0.054,为较优的估测方法。腾讯街景法的拟合线斜率为0.952,表明腾讯街景法的整体估测值偏小;而非监督分类法的拟合线斜率为1.013,则表明非监督分类法的整体估测值偏大。
The study extracted the panoramic image set from Tencent Street View,transformed the cylindrical panorama into azimuth fish-eye image through image mosaic technology,and combined image enhancement,image binarization and other digital image processing techniques to extract the canopy closure of forest in Shanghai Botanical Garden.The extraction results were compared with the canopy closure calculated by the unsupervised classification method and the canopy closure of the field surveys to verify the feasibility and accuracy of the method.The results show that the absolute error mean( 0.027) and standard deviation( 0.031) of Tencent Street View method are smaller than the unsupervised classification method,which indicates that the absolute error dispersion of Tencent Street View method is small and relatively stable.The R2 of Tencent Street View is 0.977,and the RMSE is 0.054,which indicates that Tencent Street View is a better estimation method.The slope of the fitted line of Tencent Street View is 0.952,which indicates that the overall estimated value of Tencent Street View is smaller; the slope of the fitted line of unsupervised classification method is 1.013,indicating that the overall estimated value of unsupervised classification method is larger.
The study extracted the panoramic image set from Tencent Street View,transformed the cylindrical panorama into azimuth fish-eye image through image mosaic technology,and combined image enhancement,image binarization and other digital image processing techniques to extract the canopy closure of forest in Shanghai Botanical Garden.The extraction results were compared with the canopy closure calculated by the unsupervised classification method and the canopy closure of the field surveys to verify the feasibility and accuracy of the method.The results show that the absolute error mean( 0.027) and standard deviation( 0.031) of Tencent Street View method are smaller than the unsupervised classification method,which indicates that the absolute error dispersion of Tencent Street View method is small and relatively stable.The R2 of Tencent Street View is 0.977,and the RMSE is 0.054,which indicates that Tencent Street View is a better estimation method.The slope of the fitted line of Tencent Street View is 0.952,which indicates that the overall estimated value of Tencent Street View is smaller; the slope of the fitted line of unsupervised classification method is 1.013,indicating that the overall estimated value of unsupervised classification method is larger.
引文
[1]朱教君,康宏樟,胡理乐.应用全天空照片估计林分透光孔隙度(郁闭度)[J].生态学杂志,2005,24(10):1234-1240.
[2]祁有祥,骆汉,赵廷宁.基于鱼眼镜头的林冠郁闭度简易测量方法[J].北京林业大学学报,2009,31(6):60-66.
[3]刘芳,杨广斌.基于鱼眼照片的森林郁闭度快速提取方法研究[J].西南林业大学学报,2013,33(2):71-74.
[4]任德智,葛立雯,罗大庆.小尺度森林郁闭度测定方法及应用[J].林业科技开发,2014,28(1):94-97.
[5]German D M,D'Angelo P,Gross M,et al.New methods to project panoramas for practical and aesthetic purposes[C]//Eurographics Association.Computational Aesthetics'07 Proceedings of the Third Eurographics conference on Computational Aesthetics in Graphics,Visualization and Imaging.Alberta,Canada:Eurographics Association,2007:15-22.
[6]D'Angelo P.Radiometric alignment and vignetting calibration[C]//Applied Computer Science Group,Bielefeld University.Proceedings of the ICVS Workshop on Camera Calibration Methods for Computer Vision Systems.Bielefeld,Germany:Bielefeld University,2007.
[7]Glatthorn J,Beckschfer P.Standardizing the protocol for hemispherical photographs:accuracy assessment of binarization algorithms[J].PLoS ONE,2014,9(11):e111924.
[8]Brusa A,Bunker D E.Increasing the precision of canopy closure estimates from hemispherical photography:Blue channel analysis and under-exposure[J].Agricultural and Forest Meteorology,2014,195/196:102-107.
[9]Loffredo N,Sun X,Onda Y.DHPT 1.0:New software for automatic analysis of canopy closure from under-exposed and over-exposed digital hemispherical photographs[J].Computers and Electronics in Agriculture,2016,125:39-47.
[10]Hwang R L,Lin T P,Matzarakis A.Seasonal effects of urban street shading on long-term outdoor thermal comfort[J].Building and Environment,2011,46(4):863-870.
[11]Klemm W,Heusinkveld B G,Lenzholzer S,et al.Street greenery and its physical and psychological impact on thermal comfort[J].Landscape and Urban Planning,2015,138:87-98.
[12]Huang K T,Lin T P,Lien H C.Investigating thermal comfort and user behaviors in outdoor spaces:a seasonal and spatial perspective[J].Advances in Meteorology,2015,2015:1-11.
[2]祁有祥,骆汉,赵廷宁.基于鱼眼镜头的林冠郁闭度简易测量方法[J].北京林业大学学报,2009,31(6):60-66.
[3]刘芳,杨广斌.基于鱼眼照片的森林郁闭度快速提取方法研究[J].西南林业大学学报,2013,33(2):71-74.
[4]任德智,葛立雯,罗大庆.小尺度森林郁闭度测定方法及应用[J].林业科技开发,2014,28(1):94-97.
[5]German D M,D'Angelo P,Gross M,et al.New methods to project panoramas for practical and aesthetic purposes[C]//Eurographics Association.Computational Aesthetics'07 Proceedings of the Third Eurographics conference on Computational Aesthetics in Graphics,Visualization and Imaging.Alberta,Canada:Eurographics Association,2007:15-22.
[6]D'Angelo P.Radiometric alignment and vignetting calibration[C]//Applied Computer Science Group,Bielefeld University.Proceedings of the ICVS Workshop on Camera Calibration Methods for Computer Vision Systems.Bielefeld,Germany:Bielefeld University,2007.
[7]Glatthorn J,Beckschfer P.Standardizing the protocol for hemispherical photographs:accuracy assessment of binarization algorithms[J].PLoS ONE,2014,9(11):e111924.
[8]Brusa A,Bunker D E.Increasing the precision of canopy closure estimates from hemispherical photography:Blue channel analysis and under-exposure[J].Agricultural and Forest Meteorology,2014,195/196:102-107.
[9]Loffredo N,Sun X,Onda Y.DHPT 1.0:New software for automatic analysis of canopy closure from under-exposed and over-exposed digital hemispherical photographs[J].Computers and Electronics in Agriculture,2016,125:39-47.
[10]Hwang R L,Lin T P,Matzarakis A.Seasonal effects of urban street shading on long-term outdoor thermal comfort[J].Building and Environment,2011,46(4):863-870.
[11]Klemm W,Heusinkveld B G,Lenzholzer S,et al.Street greenery and its physical and psychological impact on thermal comfort[J].Landscape and Urban Planning,2015,138:87-98.
[12]Huang K T,Lin T P,Lien H C.Investigating thermal comfort and user behaviors in outdoor spaces:a seasonal and spatial perspective[J].Advances in Meteorology,2015,2015:1-11.