| As a main factor of the ecological construction,the urban green land can improve the living environment and recreational conditions,which possesses an irreplaceable function on the aspects of economic benefit,ecological benefits and so on.However,the planning level of the urban green land has also correspondingly exposed lots of shortcomings and defects at the same time of continuous improvement.Excessive pursuit of the rapid development of economy even at the cost of destroying the ecological environment leads the city being sub-healthy.Therefore,it is of great guiding significance on related government departments planning the urban green land by applying the latest GF-2 remote-sensing image to extract and analyze the urban green land information.The utilization of the domestic GF-2 remote-sensing image in 2014 had a great significance on the extraction of green land information,the research on ecological sustainability of green land became much more completed with the support of informationalized,modern and scientific technologies.This paper will set the built up area of Xixiangtang in Nanning as an example,with the domestic GF-2 remote-sensing image of June,2016 as the main information source to conduct the radiometric calibration,RPC orthographic adjustment,FLAASH atmosphere correction and image registering on the original statistics of GF-2 remote-sensing image based on the software ENVI.According to the national green land taxonomy combined with the research land conditions,by fusing investigating methods,the urban green land were divided into 6 types and the analysis of its landscape pattern was proceeded under the help of Fragstats software so as to provide rational gist for the future urban green land planning.And here are the main results:(1)Fusing images by using 4 different methods and receiving the NND fusion technology which can keep the spectrum capability at strongest level with relatively clear spatial textural features by subjective qualitative and objective quantitative analysis.On the basis of the study about NND fusion technology on fusion speed,the fusion speed has quickened almost for 3 times when the multi-spectral format is BIP.(2)During the extraction of the green land information,we utilized the objext-oriented multiresolution segmentation method and after repeatedly experiments of the testing sample,we drew to a conclude that when the segmentation threshold value was set at 35 and the merge threshold value was set at 75,the result was the best;the object-oriented extraction of the vegetation rules were applied into:the NDVI value was 0.364,the ND value of the blue band was less than 1330,which was able to extract the vegetation information better.(3)Through analyzing the landscape indicators,we got that the attached green land's NF2 of the urban green land researched area is 0.865 and MPS is 0.042,which indicated that the fragmentation degree of the attached green landis high and the green lands are badly fragmented.While,the MPS of the public green land is relatively high,yet the aggregate area is small and uneven distributed,the street recreational green land is little as well.(4)The landscape diversification indicator of the researched area is 1.531,the evenness index is 0.712 and the average plaque fractal dimension is 1.385,hence the overall fragmentation degree is relatively high,yet the various green land area distribution is relative even.However,the green land of the remaining developed area and the rural-urban continuum are rather dispersed,the integrality is not that strong without forming a continuous ecological chain and its ecological and social functions are notably low. |
PDF Full Text Request