Dissertation > Excellent graduate degree dissertation topics show

Study on Extraction of Coniferous Forest Information in Southern China

Author: MaYanHui
Tutor: LinHui
School: Central South University of Forestry Science and Technology
Course: Forest Management
Keywords: MODIS EVI Time Series Decision Tree Information Extraction Coniferous forest
CLC: TP79
Type: Master's thesis
Year: 2010
Downloads: 149
Quote: 0
Read: Download Dissertation

Abstract


The forest is an important part of the land cover are important indicators of the ecological environment of the region, is an important material basis of the national strategy of sustainable development. Access to forest resources status and its changes have important significance of forest management decisions. Although the traditional forest resources survey to accurately obtain the distribution of forest resources, but consuming huge human, material and financial resources, and lasted longer, the findings are often subjective. The development of remote sensing technology provides a new means for the study of regional forest resources situation the MODIS data because of its band rich, real-time imaging range, has been widely used in the monitoring of forest resources. Research coniferous forest using MODIS data extraction, aimed at monitoring of forest resources, protection, the survey provides reference data for the relevant departments to provide a scientific basis for decision-making. Papers in the \time series MOD13Q1 data coniferous forest in the remote sensing information extraction image pre-processing, the best band combination selection NDVI and EVI reaction forest vegetation seasonal variation in sensitivity, maximum likelihood and support vector machine remote sensing image classification and knowledge-based rules The decision tree classification techniques to carry out a more systematic study. The main findings are as follows: 1) the MOD09A1 data 7 band spectrum band 2 to carry the feature the largest amount of information between the different surface features greater separability MODIS applied research focused on the band. 2) through the the MOD09A1 data seven original band on the comparative analysis of different forest types in the spectral characteristics of different forest types in each band spectral response curve trends approximation, and the close proximity of the use of single-phase MODIS data is difficult to distinguish between forest types . But after the stretch after 2-6-3 band composite image of bamboo ability to identify significantly enhanced. 3) the best index, the MOD09A1 data theories of the study area the best RGB synthesis 2-5-3 band found between 2-3 -6-3 band synthetic images of different objects quite different spectral characteristics, but combined with the specific circumstances of advantageous identification coverings. 4) found by comparative analysis of two kinds of MODIS vegetation index NDVI and EVI changes the EVI can better reaction season of different forest types (evergreen coniferous forest, evergreen broadleaf forest, deciduous broadleaf forests, bamboo groves and bushes) variations. 5) by a single-phase the MOD09A1 data, band 2, band only 3, band 6, band group EVI, NDSI build the MLC and SVM classification, two classification methods Comparatively speaking, the SVM method overall accuracy than MLC, but two kinds of methods are difficult to distinguish between the different forest types. 6) for the time series EVI data, the overall accuracy of the decision tree classification algorithm reached 74.22%, the Kappa coefficient of 0.6984, the coniferous forest user accuracy to 79.13%. Multitemporal MODIS data extracted according to the law of the growth of vegetation, forest type information. Decision tree classifier to achieve the different spatial resolution of the mixed-use, multi-source data between the data and can easily take advantage of a variety of geographic information, the combination of GIS data and RS images can be satisfied with the results of the classification. Discrimination rules in the decision tree, the introduction of the DEM, slope data, developed a more scientific classification accuracy, especially in mountainous and hilly areas. Should be maximized in the latter part of the research work in multi-source information to aid in the remote sensing information extraction.

Related Dissertations

  1. Research on Domain Entity Attribute and Event Extraction Technology,TP391.1
  2. Research on Temporal Information Recognition and Normalization,TP391.1
  3. Fault Diagnosis Method Based on Support Vector Machine,TP18
  4. Study on the Retrieval of Aerosol Optical Thickness Based on HJ and MODIS Data,X513
  5. Study on Growth Monitoring Technique Based on Pixel Un-Mixing Method and HJ Remote Sensing Images in Paddy Rice,S511
  6. Study on Method of Improving Cultivated Area Extraction from Modis Image Based on Scale Transfromation,S127
  7. Crop Evapotranspiration Study on Evolution Rule and Forecast Model in Chaoyang Area,S161.4
  8. Three Rivers District quantitative estimation of vegetation cover and its changing,TP79
  9. Active faults based radar image information extraction method applied research and demonstration,P542.3
  10. Based on Data Mining Technologies in Urban Water Supply Analysis and Decision,F299.24;F224
  11. Web Page Attribute Extraction Method Research,TP391.1
  12. Home Academic Information Extraction System,TP393.092
  13. Topic search engine key technology research,TP391.3
  14. Hull section robotic welding path planning and offline programming,TP242
  15. Based on semi- structured text transporter protein substrate information extraction system,Q811.4
  16. Decision Tree Based on Generalized Information Entropy Model and Its Performance Evaluation,TP18
  17. Dynamic learning framework based on structured automatic web data extraction method,TP393.092
  18. Printers based on natural language HCI Research and implementation,TP11
  19. Short-term load forecasting technology,TM715
  20. Research on the Subway Travel Demand Based on Multiply Social & Economic Factors,U231
  21. Characteristics Analysis for DNA Sequences of the Influenza Virus Based on the Time Series Theory Methods,R346

CLC: > Industrial Technology > Automation technology,computer technology > Remote sensing technology > The application of remote sensing technology
© 2012 www.DissertationTopic.Net  Mobile