| Along with the rapid development of social economy,precision poverty alleviation and rural revitalization are being implemented in full,and the spatial pattern of settlements is being changed.Fast and accurate acquisition of settlement information and its optimal evaluation have both theoretical and practical significance.Taking Luojiang District of Deyang City as an example,based on two remote sensing images of GF-1 and Landsat-8,this paper employs object-oriented classification method to extract the settlement information of Luojiang District,on the basis of which,the spatial distribution pattern of settlements and the factors affecting the spatial distribution of settlements are analyzed by using GIS spatial technology,landscape index method,Geo-detectors,etc.At last,AHP and GIS weighted index summation method are used to evaluate the suitability of settlements in Luojiang District,and scientific suggestions are put forward for the sustainable development and optimization of settlements.The results of the study are as follows:(1)By analyzing the spectral features,texture features,spatial geometry features and derived data features of different ground objects,the models based on knowledge were established,and used to extract the settlements from images.The accuracy of the settlement extraction was evaluated by using the confusion matrix and kappa coefficient.The overall accuracy of the settlement extraction was respectively 91.1% for GF-1 and77.63% for Landsat-8.The results of settlement extraction of GF-1 remote sensing images are ideal.After the results were improved by correcting some errors,they were used for the further research.(2)The settlement of Luojiang District is mainly distributed in the area with lower elevation,lower slope and closer to road and river.The dominant areas of settlement distribution are Jinshan Town and Wan’an Town,but their internal order is poor,which needs further planning.The patch shape of settlement in Wanan Town,Yuying Town is relatively simple,high degree of agglomeration,and landscape fragmentation is low,but the patch shape of settlement in Panlong Town,Xinsheng Town,Yanjia Town is high degree of fragmentation and dispersion.(3)Via the use of Geo-detectors to detect the influencing factors of the settlement distribution in Luojiang District,the distance of villages and towns,the per capital cultivated land,the per capital disposable income and the population density are the major factor of distribution of settlement in Luojiang District.The settlement distribution in Luojiang District is affected by multiple factors instead of a single factor,which is obvious by interaction between two factors.(4)The suitability evaluation and optimization results of Luojiang settlement show that Luojiang settlement can be divided into four types: core development,transitional development,adjustment development and restricted development,of which the core development type has the largest area,accounting for 49.6% of the total settlement area,and the settlement area of restricted development type is the smallest,accounting for2.15% of the total settlement area.(5)Suggestions on settlement optimization in Luojiang District: In terms of the core development type settlements,it should play the role of central cities and towns,attract scattered settlements around,realize resource sharing,absorb foreign capital and enhance economic vitality;With regards to the transitional development type settlement,it should actively move closer to the core development type settlement and play a good bridge role between the core region and the adjustment region;Regarding the adjustment development type settlements,it should incorporate remote scattered settlements into the regions with convenient transportation,rebuild the dangerous buildings,improve the quality of the ecological environment;Regarding the restricted development type settlements,attention should be paid to the protection of the ecological environment.Under the implementation of the precise poverty alleviation policy,remote scattered settlements should be relocated and dilapidated buildings should be renovated. |
