The Planting Information Investigation And System Development For Korla Pears Based On3S Technology

Fragrant pear business is one of the most important pillar industries offorestry in Korla. For mastering the planting and production status of Korla pearsin time, enhancing administrative decision-making ability and meet the need ofmodern information management, the investigation and information management forKorla pears have been carried on based on3S Technology according to the demandof Korla municipal government.Taking administrative region of Korla as study area and the latest ZY-3satellite remote-sensing image as basic information source, the remote sensinginvestigation and information management for Korla pears are carried on using thefield survey data obtained by GPS and Google Earth. The main study results areas follow:1.By collecting ZY-3satellite remote-sensing image data of researched area,adopting professional image processing system, the image data is preprocessed bygeometric correction with precision, band rectification, inlay and RGB false colorcomposite, which is used for the basic data image of remote-sensing interpretationfor Korla pears.2. Analyzing the spectral and texture features of image data, the extractingmodel of decision-making tree for Korla pears has been constructed based on thespectral feature and spectrum combined with texture feature. Using therepresentative part of ZY-3satellite remote-sensing images as test site, theplanting information in test site was extracted respectively by two models. Thenwe analyzed the effect of extraction and selected the better from the two modelsas the extraction model. The planting area and spatial distributed informationcan be obtained by extraction (the preliminary survey achievements are used inthe paper).3. It can be found by means of precision test that the overall accuracy ofKorla pears extraction reaches93.1%, Kappa coefficient is0.86, the accuracy ofmap-making for pear is93.62%, and the omission error is6.38%. The identificationaccuracy for the ages of Korla pears is86.36%and Kappa coefficient is0.79.4. Using the C#as development language, the.NET as development environment,ArcGIS Server as development platform, the information system for Korla pear planting is developed by combining SQL SERVER2005database technology and ArcSDEspatial database engine, which can achieve the basic functions such as browse,map query, data maintenance, planting information statistics and thematic mapmaking. The system can provide real-time, reliable and fast scientific decisionsupport for relevant departments or decision makers and improve management levelfor Korla pear industry.