Study On Automated Selection Of Map Elements Based On Intelligent Algorithm

The map,which was the most important medium for storing and transmitting of spatial information,was an abstract representation of the objective world.And it had play an important role in human production and life.Especially today,with the increasingly mature of geographic information technology,map had became an indispensable tool in all aspects,such as daily travel,agricultural production,transportation,and urban planning.In order to meet the needs of human activities,more and more map scales were required.At the same time,with the rapid development of human society,map elements have changed more and more quickly,which required that map had to update faster.If we made different scale maps with high timeliness,we had to invest a lot of manpower,material resources and financial resources.Therefore,studying automated map generalization was particularly important.Automated map generalization referred to make the computer transform different scales of maps by selecting,simplifying,and displacing operators.In the paper,we mainly studied the automated selection of point elements and linear elements.The main contents of this paper include the following aspects:(1)Taking settlement as an example,we had realize the automated selection of point elements.In this paper,the settlements in haidian district of Beijing included population density data and area date were used as experimental data.Firstly,we used convex hull algorithm to divide the points into inner points and outer points.As for outer points,the Douglas-Peucker algorithm was used to realize the selection operation.And as for inner points,we used the self-organizing maps algorithm to realize the selection operation.Then these two results were combined to form the selection result of the final points.In addition,We evaluated the quality of the results in three aspects: statistical information,thematic information and measurement information.(2)Taking river as an example,we realized the automated selection of linear elements.In this paper,the Danjiang River network with a scale of 1:150 million was used as the experimental data.Firstly,river entity and river tree were constructed by river source.At the same time,the length,level,river code of each river were calculated.Then,the paper designed an affinity function which could take into account the above three indexes simultaneously.In the end,we used immune algorithm to get the optimal solution and evaluated the quality of results by comparing the results of manual selection.The conclusions of this paper are as follows:(1)The automated selection method of point elements,which add the population density attribute and residential area attribute as the weighted,was more in line with practical requirements.(2)Self-organizing Maps algorithm could well determine which points generate “competition”.So it is suitable for automated selection of point elements.(3)we transformed the problem that automated selection of linear elements to an optimization problem,the immune algorithm was very suitable for finding this optimal solution.