Urban green space Information extraction scheme with high resolution image supported by ENVI

Urban green space plays an active role in improving urban ecological environment and human settlement environment, and the content of urban green space is becoming an important index to measure the quality of urban life. In addition, the spatial distribution pattern of urban green space is closely related to its ecological effect. Therefore, it is necessary to master the information of urban Greenland objectively and accurately. The traditional urban greening survey was completed by the basic unit Reporting statistics and the field Sampling survey, and the capital and manpower input was big and the time period was long. The data is influenced by man and has low precision, and lacks the function of spatial statistic analysis.

with space remote sensing technology development

at present, commercially available satellite imagery can be obtained up to 0.5 meters, can distinguish the common road in the middle of the green belt, even single trees. and has high spectral resolution, such as including Red Band, near infrared band, for accurate automatic extraction of urban green space information provides a prerequisite. A high-resolution image can cover The range of 18.5kmx18.5km,2-3 Day can be repeated shooting in the same area, can be carried out in a wide range, short cycle of investigation.

Technical Process

As a typical urban green space information extraction process based on high resolution image, it involves high resolution image rectification, image fusion, atmospheric correction, object-oriented image information extraction, vector editing and processing, attribute assignment and so on.

in addition to usingENVIIn addition to the main module function, a fast atmospheric correction tool in the Atmospheric Correction expansion module (QuAC),ENVI EXin the extension moduleFeature ExtractionTools,arcgis@ Desktopof theArcMAP.

Information extraction process of urban green space based on high resolution image

Key Technologies

First, Data Acquisition

select with rpc file data, Data including multispectral and panchromatic bands, imaging time 6~9 dem data, control point data or control point selection source.

Ii. data preprocessing

According to the characteristics of high-resolution satellite imagery now, The fusion of panchromatic and multispectral images is done first, and then the rpc file, and the accuracy of the positive shot correction results of the fused images is consistent with that of the panchromatic image. This sequence reduces the process and increases efficiency, and enables accurate spatial registration between panchromatic and multispectral image fusion. Use envi pansharpening Fusion method, which is specifically designed for high-resolution imagery.

based on control point +rpc+dem Complete the ortho correction process, Control points are selected from the reference image, or they can be obtained using field measurements, and the number of control points for each image should not be too large ( 8~12

Use the Fast Atmospheric correction tool ( quac ) Remove the influence of some atmospheres, In the process of object-oriented green information extraction, the ndvi

Iii.

using envi The Face object tool ( feature Extraction " To do this, the tool is guided, easy to use, and has a real-time preview function.

Because the high-resolution image data is generally large, in order to get the rules quickly. Select a part of the area as the study area to determine the object segmentation and merging thresholds, rule-based information extraction of the object thresholds, and then the experimental area to obtain the thresholds and rules applied to the entire image file. Sample statistics can be used.

Flowchart of Sample Statistics method

Four, green space vector result processing

   The whole process is inarcgis@ Desktopof theArcMAP, including vector results checking and editing, vector data stitching and cropping, and attribute assignment. Feature Extractiontool can choose to output the vector results asShapefilesformat orArcGIS geodatabaseformat, all areArcGISvector format files.

The important part of this step is attribute assignment, the green space vector result includes area attribute field, also need to increase the Greenland type field. If the obtained green space vector results are divided into: Park green space, production green space, protective green space, ancillary green space, other green space. Completing this process requires a vector of data: Urban land classification. UseArcMAPIdentification tool for two vector data identification analysis, the "Urban Land Classification" and "green space vector" spatial map to add urban land classification information, reference to the People's Republic of China industry Standards "urban Greenland classification standards-cjj/t 85-2002further attribute assignment. ArcMAPThe identification analysis tools in:Arctoolbox->analyst tools->overlay->identity.

Use WorldView-2 Video & the results of urban green space information extraction based on object-oriented taxonomy

Urban green space Information extraction scheme with high resolution image supported by ENVI