Optimum Temporal And Remote Sensing Extraction Of Land Use/Cover Information In The Yellow River Delta

The land is indispensable to the survival of human resources, so we must understand the situation of land resources to achieve the sustainable use of land resources. Remote sensing is an important monitoring technique of resources survey, environmental monitoring and land use/cover information extraction and dynamic. Academia has always been key areas of concern on the Yellow River Delta due to the formation of special features, and typical of the ecological environment and vulnerability. As the connection area of the river and marine systems, the area of land type is complex and diverse, serious soil salinization, poor quality of cultivated land and arable land, saline-alkali soil, Grassland and Forest there is a flower arrangement distribution to class mixed spectral characteristics, construction land and other small areadifficult to directly extract. As the connection area of the river and marine systems, the area of land type is complex and diverse, including poor quality of cultivated land and arable land, saline-alkali soil, Grassland and Forest there is a flower arrangement distribution to class mixed spectral characteristics, construction land and other small area, so it is difficult to directly extract serious soil salinization.In recent years, the socio-economic development and natural resources conflicts are deepening, soil and environment quality is severely affected with the population growing, including the destruction of the ecosystem of the Yellow River Delta, shrinking wetlands, soil salinization, water stress andserious pollution, diminished vegetation, biological drastic reduction in the number of serious environmental problems, which restricts social and economic development in the area.Under the double impact of natural factors and human factors, the Yellow River Delta land use structure has undergone great changes, affecting the healthy development of the area of land resources. In previous research work, the common practice is to use the area of a relative or several phase of remote sensing data processing to extract the land use/cover change information. However, there is less straddles the study of the short period of the year from the macro range. Therefore, this article is intended to be Yellow River Delta kenli County as study area, based on TM remote sensing data.The classification accuracy of the results compared. Yellow River Delta over the past year land use/cover classification, exploring the area to extract the best phase of a class and methods, which is good for natural resources to provide scientific reference for the protection of the ecological environment of the Yellow River Delta region,and the rational and sustainable use of regional land.This article is intended to be Yellow River Delta kenli County as study area,there is overview of remote sensing technology in land use/cover research status in the information research firstly, followed by the analysis of the natural and socio-economic profile of the study area, and analysis of remote sensing technology in land use/cover remote sensing extraction study the status quo. It uses a supervised classification, object-oriented and knowledge-based three classification methods for data processing respectively with March2009and June, September2010and January TM remote sensing data based in the ENVI software support, and extract the main feature of the study area of arable land, forest and grassland, salt wasteland, waters, tidal flats and land for construction on the basis of these data. The purpose of the article is to get parative analysis and study area in spring, summer, autumn and winter seasons of different objects land use/cover classification accuracy of the information using Kenli County2009land-use status map, Statistical Yearbook and soil nutrient content map. The main research results and content:(1) Kenli County seasons is selected a total of four times four remote sensing images, geometric correction, image fusion, image enhancement, image cropping, processing and information extraction, based on the existing land use classification system and the use of remote sensing data, combined with the land of the study areause/cover characteristics, to establish a unified classification system, land type of the study area is divided into the following six types:waters (including reservoirs and rivers), forest and grassland, construction land, salt wasteland, arable land (including the dry land and paddy fields), beaches.(2) Remote sensing is an important means of monitoring land use, the use of multi-temporal remote sensing data can be well to distinguish between land use/cover types, four remote sensing image analysis, remote sensing image derived using the March and June can be more accurateextraction of land use/cover types.(3) There is the main use of supervised classification, object-oriented and knowledge-based classification three methods for image classification by the Kenli of the Yellow River Delta four remote sensing image processing, based on the analysis of the spectral characteristics of the class around. On the basis of the four remote sensing image processing, the use of supervised classification, object-oriented and knowledge-based classification to class information is extracted, the highest image in March overall accuracy and kappa coefficient, and the lowest in January for coastal areas land use classification and changes in remote sensing monitoring provides effective technical support.(4) The results by contrast March, June, September, January four image classification accuracy show that the use of overall accuracy of the classification of knowledge-based and Kappa coefficient of the highest overall accuracy were91.43%,89.71%,88.80%,79.71%, the Kappa coefficient0.8931,0.8649,0.8478and0.7239.(5) Comprehensive analysis of four remote sensing producer accuracy of the image classification process, the user accuracy, overall accuracy and Kappa coefficient shows that the June images with knowledge-based classification method to extract the most suitable waters and land for construction, March image based on knowledgeextraction of forest and grassland, salt wasteland, arable land and beach is the most appropriate. Therefore, a knowledge-based approach to land use/cover classification, March is the most suitable phase.(6) Image processing on September is confused due to the spectral information of the arable land and saline wasteland very serious; almost impossible to distinguish it will merge the two. Object-oriented approach to processing the images of September of arable land and salt wasteland combined classification accuracy can be higher, but both difficult to distinguish, it can not be selected in September as to extract the most appropriate of the arable land and salt wasteland phase.