针对SLIC(Simple Linear Iterative Clustering)算法在超像素聚类过程中耗时较长的缺陷,提出一种基于快速SLIC的图像超像素算法。该算法首先剔除在颜色空间上与聚类中心相似度较低的像素,从而仅用部分近邻像素更新聚类中心,以确保聚类...针对SLIC(Simple Linear Iterative Clustering)算法在超像素聚类过程中耗时较长的缺陷,提出一种基于快速SLIC的图像超像素算法。该算法首先剔除在颜色空间上与聚类中心相似度较低的像素,从而仅用部分近邻像素更新聚类中心,以确保聚类中心快速达到稳定并阻止误差传播,提高边缘命中率;其次,在初始化网格后,将每个超像素的边缘像素视为不稳定像素,将超像素的非边缘像素视为稳定像素并保持稳定像素的类别不变;最后,通过对不稳定像素进行迭代标记来实现快速超像素图像分割。在MATLAB环境下分别对所提算法与6种对比算法进行测试,在超像素个数相同的情况下,所提算法在BSD500数据集上与经典的SLIC算法相比分割误差率降低5%,分割精度提高0.5%,运行时间减少0.18 s。实验结果表明,与主流的超像素算法相比,所提算法在提升超像素分割质量的同时能够有效降低算法的计算复杂度。展开更多
The spatial variability in the concentrations of 1,2,3,4,5,6-hexachlorocyclohexane （HCH） and 1,1,1-trichloro-2,2-bis-（p-chloro- phenyl） ethane （DDT） in surface soils was studied on the basis of the analysis of 1...The spatial variability in the concentrations of 1,2,3,4,5,6-hexachlorocyclohexane （HCH） and 1,1,1-trichloro-2,2-bis-（p-chloro- phenyl） ethane （DDT） in surface soils was studied on the basis of the analysis of 131 soil samples collected from the surface layer （0-20 cm depth） of the alluvial region of Beijing, China. The concentrations of total HCHs （including α-,β-, γ-, and δ-isomers） and total DDTs （including p,p＇-DDT, p,p＇-DDD, p,p＇-DDE, and o,p＇-DDT） in the surface soils tested were in the range from nondetectable to 31.72 μg/kg dry soil, with a mean value of 0.91, and from nondetectable to 5910.83 μg/kg dry soil, with a mean value of 32.13, respectively. It was observed that concentrations of HCHs in all soil samples and concentrations of DDTs in 112 soil samples were much lower than the first grade （50 μg/kg） permitted in ＂Environment quality standard for soils in China （GB 15618-1995）＂. This suggests that the pollution due to organochlorine pesticides was generally not significant in the farmland soils in the Beijing alluvial region. In this study, the spatial distribution and trend of HCHs and DDTs were analyzed using Geostatistical Analyst and GS＋（513）. Spatial distribution indicated how these pesticides had been applied in the past. Trend analysis showed that the concentrations of HCHs, DDTs, and their related metabolites followed an obvious distribution trend in the surface soils from the alluvial region of Beijing.展开更多
文摘针对SLIC(Simple Linear Iterative Clustering)算法在超像素聚类过程中耗时较长的缺陷,提出一种基于快速SLIC的图像超像素算法。该算法首先剔除在颜色空间上与聚类中心相似度较低的像素,从而仅用部分近邻像素更新聚类中心,以确保聚类中心快速达到稳定并阻止误差传播,提高边缘命中率;其次,在初始化网格后,将每个超像素的边缘像素视为不稳定像素,将超像素的非边缘像素视为稳定像素并保持稳定像素的类别不变;最后,通过对不稳定像素进行迭代标记来实现快速超像素图像分割。在MATLAB环境下分别对所提算法与6种对比算法进行测试,在超像素个数相同的情况下,所提算法在BSD500数据集上与经典的SLIC算法相比分割误差率降低5%,分割精度提高0.5%,运行时间减少0.18 s。实验结果表明,与主流的超像素算法相比,所提算法在提升超像素分割质量的同时能够有效降低算法的计算复杂度。
基金Project supported by the National Natural Science Foundation of China (No. 40201023)the Program for Changjiang Scholars and Innovative Research Team in University (IRT0412).
文摘The spatial variability in the concentrations of 1,2,3,4,5,6-hexachlorocyclohexane （HCH） and 1,1,1-trichloro-2,2-bis-（p-chloro- phenyl） ethane （DDT） in surface soils was studied on the basis of the analysis of 131 soil samples collected from the surface layer （0-20 cm depth） of the alluvial region of Beijing, China. The concentrations of total HCHs （including α-,β-, γ-, and δ-isomers） and total DDTs （including p,p＇-DDT, p,p＇-DDD, p,p＇-DDE, and o,p＇-DDT） in the surface soils tested were in the range from nondetectable to 31.72 μg/kg dry soil, with a mean value of 0.91, and from nondetectable to 5910.83 μg/kg dry soil, with a mean value of 32.13, respectively. It was observed that concentrations of HCHs in all soil samples and concentrations of DDTs in 112 soil samples were much lower than the first grade （50 μg/kg） permitted in ＂Environment quality standard for soils in China （GB 15618-1995）＂. This suggests that the pollution due to organochlorine pesticides was generally not significant in the farmland soils in the Beijing alluvial region. In this study, the spatial distribution and trend of HCHs and DDTs were analyzed using Geostatistical Analyst and GS＋（513）. Spatial distribution indicated how these pesticides had been applied in the past. Trend analysis showed that the concentrations of HCHs, DDTs, and their related metabolites followed an obvious distribution trend in the surface soils from the alluvial region of Beijing.
