为了提高车道线检测的准确性与鲁棒性,降低光照变化与背景干扰的影响,提出了一种改进的Hough变换耦合密度空间聚类的车道线检测算法。首先,建立车道线模型,将车道边界分解为一系列的小线段,借助最小二乘法来表示车道线中的线段。再利用...为了提高车道线检测的准确性与鲁棒性,降低光照变化与背景干扰的影响,提出了一种改进的Hough变换耦合密度空间聚类的车道线检测算法。首先,建立车道线模型,将车道边界分解为一系列的小线段,借助最小二乘法来表示车道线中的线段。再利用改进的Hough变换对图像中的小线段进行检测。引入具有密度空间聚类方法(density based spatial clustering of applications with noise,DBSCAN),对提取的小线段进行聚类,过滤掉图像中的冗余和噪声,同时保留车道边界的关键信息。随后,利用边缘像素的梯度方向来定义小线段的方向,使得边界同一侧的小线段具有相同的方向,而位于相反车道边界的两个小线段具有相反的方向,通过小线段的方向函数得到车道线段候选簇。最后,根据得到的小线段候选簇,利用消失点来拟合最终车道线。在Caltech数据集与实际道路中进行测试,数据表明:与当前流行的车道线检测算法相比,在光照变化、背景干扰等不良因素下,所以算法呈现出更理想的准确性与稳健,可准确识别正常车道线。展开更多
针对混合矩阵估计算法中传统的噪声环境下基于密度的空间聚类(density-based spatial clustering of applications with noise,DBSCAN)算法需要人为设定邻域半径以及核心点数这一问题,提出双约束粒子群优化(double constrained particle...针对混合矩阵估计算法中传统的噪声环境下基于密度的空间聚类(density-based spatial clustering of applications with noise,DBSCAN)算法需要人为设定邻域半径以及核心点数这一问题,提出双约束粒子群优化(double constrained particle swarm optimization,DCPSO)算法,对DBSCAN算法的邻域半径参数进行寻优,将得到的最优参数作为DBSCAN算法的参数输入,然后计算聚类中心,完成混合矩阵估计。针对基于距离排序的源信号数目估计算法存在依靠经验参数的选取且不具备噪声点剔除能力的问题,提出了最大距离排序算法。实验结果表明,所提算法较相应的对比算法皆有提升,源信号数目估计准确率较原算法提高近40%,混合矩阵估计的误差较对比算法提升3 dB以上,且所提算法在收敛速度上优于原算法。展开更多
数字经济时代POI大数据为城市及城市群商业空间结构及变化研究提供了新思路。基于2015年和2021年POI大数据,采用空间核密度、Theil指数等方法从区县层面对粤港澳大湾区商业总体及购物、休闲、餐饮三类业态的空间布局演变特征、区域差异...数字经济时代POI大数据为城市及城市群商业空间结构及变化研究提供了新思路。基于2015年和2021年POI大数据,采用空间核密度、Theil指数等方法从区县层面对粤港澳大湾区商业总体及购物、休闲、餐饮三类业态的空间布局演变特征、区域差异进行研究。结果表明:1) 2015~2021年间粤港澳大湾区多中心商业空间一体化发展格局进一步强化,形成显著的多中心多等级都市圈商业空间格局;2) Theil指数表明商业网点总体及三大细分业态POI数量的区域差异均呈现扩大态势,组内差距明显大于组间差距;相对于城市尺度下的区域差异,区县尺度下的组内差异有所下降但组间差距明显增大;3) 不同商业中心区的演变趋势存在差异,高密度商圈主要分布于广州、深圳、香港三大一线城市,商业网点密度最高等级从2015年的968个/km2增加到2021年1904个/km2;4) 购物服务类、餐饮服务类、休闲娱乐POI增长幅度均超过50%,空间集聚和连片化特征明显加强;5) 不同商业业态网点规模结构发生动态调整,超市、专卖店、便利店等业态网点增长较快,大型购物中心下降态势明显,不同业态的空间变化特征有明显差异。POI big data provides new ideas for research on spatial structure and changes of commercial space in cities and urban agglomerations in digital economy era. Based on the two periods of POI big data in 2015 and 2021, the paper used method called spatial kernel density and Theil’s Index to study evolution characteristics of spatial layout of the overall business and three types of business formats in the Guangdong-Hong Kong-Macao Greater Bay Area from perspective of county-level regions. The results showed that: 1) During the period from 2015 to 2021, the integrated development pattern of multi-center commercial space in Guangdong-Hong Kong-Macao Greater Bay Area was further strengthened, forming a significant multi-center and multi-level metropolitan commercial space pattern;2) Theil’s index showed that the regional differences of POI quantity in all commercial network and three sub-formats were enlarged, and the intra-group differences were obviously larger than the inter-group differences. Relative to regional differences at the city scale, intra-group differences at county scale declined but inter-group gaps increased significantly. 3) There were differences in the evolution trends of different commercial central areas. The high-density business districts were mainly distributed in three first-tier cities including Guangzhou, Shenzhen, and Hong Kong. The highest level of commercial network density increased from 968 per km2 in 2015 to 1904 per km2 in 2021;4) The growth rate of shopping service, catering service, and leisure entertainment POIs exceeded 50%, and the characteristics of spatial agglomeration and contiguousness were significantly strengthened;5) The scale structure of outlets for different business formats underwent dynamic adjustments. Supermarkets, specialty stores, convenience stores, and other business formats had rapid growth. Large shopping malls showed a significant downward trend. There were significant differences in the spatial variation characteristics of different formats.展开更多
文摘为了提高车道线检测的准确性与鲁棒性,降低光照变化与背景干扰的影响,提出了一种改进的Hough变换耦合密度空间聚类的车道线检测算法。首先,建立车道线模型,将车道边界分解为一系列的小线段,借助最小二乘法来表示车道线中的线段。再利用改进的Hough变换对图像中的小线段进行检测。引入具有密度空间聚类方法(density based spatial clustering of applications with noise,DBSCAN),对提取的小线段进行聚类,过滤掉图像中的冗余和噪声,同时保留车道边界的关键信息。随后,利用边缘像素的梯度方向来定义小线段的方向,使得边界同一侧的小线段具有相同的方向,而位于相反车道边界的两个小线段具有相反的方向,通过小线段的方向函数得到车道线段候选簇。最后,根据得到的小线段候选簇,利用消失点来拟合最终车道线。在Caltech数据集与实际道路中进行测试,数据表明:与当前流行的车道线检测算法相比,在光照变化、背景干扰等不良因素下,所以算法呈现出更理想的准确性与稳健,可准确识别正常车道线。
文摘数字经济时代POI大数据为城市及城市群商业空间结构及变化研究提供了新思路。基于2015年和2021年POI大数据,采用空间核密度、Theil指数等方法从区县层面对粤港澳大湾区商业总体及购物、休闲、餐饮三类业态的空间布局演变特征、区域差异进行研究。结果表明:1) 2015~2021年间粤港澳大湾区多中心商业空间一体化发展格局进一步强化,形成显著的多中心多等级都市圈商业空间格局;2) Theil指数表明商业网点总体及三大细分业态POI数量的区域差异均呈现扩大态势,组内差距明显大于组间差距;相对于城市尺度下的区域差异,区县尺度下的组内差异有所下降但组间差距明显增大;3) 不同商业中心区的演变趋势存在差异,高密度商圈主要分布于广州、深圳、香港三大一线城市,商业网点密度最高等级从2015年的968个/km2增加到2021年1904个/km2;4) 购物服务类、餐饮服务类、休闲娱乐POI增长幅度均超过50%,空间集聚和连片化特征明显加强;5) 不同商业业态网点规模结构发生动态调整,超市、专卖店、便利店等业态网点增长较快,大型购物中心下降态势明显,不同业态的空间变化特征有明显差异。POI big data provides new ideas for research on spatial structure and changes of commercial space in cities and urban agglomerations in digital economy era. Based on the two periods of POI big data in 2015 and 2021, the paper used method called spatial kernel density and Theil’s Index to study evolution characteristics of spatial layout of the overall business and three types of business formats in the Guangdong-Hong Kong-Macao Greater Bay Area from perspective of county-level regions. The results showed that: 1) During the period from 2015 to 2021, the integrated development pattern of multi-center commercial space in Guangdong-Hong Kong-Macao Greater Bay Area was further strengthened, forming a significant multi-center and multi-level metropolitan commercial space pattern;2) Theil’s index showed that the regional differences of POI quantity in all commercial network and three sub-formats were enlarged, and the intra-group differences were obviously larger than the inter-group differences. Relative to regional differences at the city scale, intra-group differences at county scale declined but inter-group gaps increased significantly. 3) There were differences in the evolution trends of different commercial central areas. The high-density business districts were mainly distributed in three first-tier cities including Guangzhou, Shenzhen, and Hong Kong. The highest level of commercial network density increased from 968 per km2 in 2015 to 1904 per km2 in 2021;4) The growth rate of shopping service, catering service, and leisure entertainment POIs exceeded 50%, and the characteristics of spatial agglomeration and contiguousness were significantly strengthened;5) The scale structure of outlets for different business formats underwent dynamic adjustments. Supermarkets, specialty stores, convenience stores, and other business formats had rapid growth. Large shopping malls showed a significant downward trend. There were significant differences in the spatial variation characteristics of different formats.
