Dissertation > Excellent graduate degree dissertation topics show

Research about the Algorithms of the Remote Image Geometry Correction Based on Support Vector Machines

Author: SunJun
Tutor: ZuoXiangYang
School: Xi'an University of Science and Technology
Course: Applied Mathematics
Keywords: Remote image Geometric correction Support vector machine (SVM) Least squares method (LSM) Correction precision
CLC: TP751;O241.5
Type: Master's thesis
Year: 2011
Downloads: 179
Quote: 2
Read: Download Dissertation

Abstract


The remote image has become an importance resource of ground-oriented observation, along with the development of space technology and remote sensing technology. But the raw remote image resources cannot be used until its varieties of distortions have been processed by the precise correction, thus to study faster and higher precision correction algorithms has become meaningful and necessary. Because the parameters of satellite orbits and sensors are not easy to obtain, here we adapt algorithms of approximate geometric correction in the remote image correcting.The current algorithms of the remote image approximate geometric correction are mainly based on the least squares method (LSM) in linear or nonlinear models. its work principle is the empirical risk minimization (ERM), the LSM has such disadvantages as over fitting, poor generalizing ability, the ill-conditioned problems in normal equation, nonsingular matrix demand, optimization solution depending on the choice of initialization solution, enough amount samples demand in order to obtain reliable solution. Thus we will introduce the theory and means of Support Vector Machine (SVM), accompanied by the essence theory of remote image approximate geometric correction, and bring forth the remote image approximate geometric correction based on SVM.In this article, we select testing region: the ground control points coordinates and altitudes are surveyed by Differential GPS, the coordinates of the ground control points in remote sensing image are measured by image processing software. By using the cluster algorithm, a fixed amount of control points are selected to act as correcting points in geometrical correcting of remote images while the rest of the control points serve as testing points. we will simultaneously apply the traditional approximate geometric correcting algorithm and the SVM algorithm, and then analyze the output of that compare the precision of both of them.The test results shows: The geometric correcting algorithm of the remote images based on Support Vector Machine, being a new approximately geometric correcting algorithm, has its feature lay in good correction precision and generalizing ability.

Related Dissertations

  1. Research on Autamatic Music Structrue Analysis,TN912.3
  2. Study on the Road Condition Monitoring Based on Vehicular 3D Acceleration Sensor,TP274
  3. Unknown orbital parameters imagery geometric correction model of,O242.1
  4. Classification Algorithm Based on Logistic Regression and Support Vector Machine,TP301.6
  5. Comprehensive Evaluation and Bottleneck Problem Analysis for the Traffic Safety of Freeway,U492.8
  6. Molecular Design Research on Aurora Kinase Target,R914.2
  7. Based on Gesture Recognition by Used Mobile Phone Universal Control,TP391.41
  8. Badan Jaran Desert dunes move some typical regional remote sensing analysis,P931.3
  9. The Drosophila wings sound research based on pattern recognition methods,TP391.41
  10. Geometric Attack Resistant Digital Image Watermarking,TP309.7
  11. Study on Fast Geometric Correction for High Resolution Satellite Image,P237
  12. Digital Watermarking Technology Based on Support Vector Machine,TP309.7
  13. Research of Face Recognition Technology,TP391.41
  14. A Dome projection system design and implementation,TP391.41
  15. Study on Scene Matching Navigation Technology Based on Real Time Image and Reference Image,V249.3
  16. Adaptive Background Model gait recognition,TP391.41
  17. A Study on Quality Control of Molding Sand,TG221
  18. Automatic Incident Detection Based on Vedio,TP274.4
  19. Study on the Hyperspectral Image Preprocessing,P237
  20. Modulation Identification for OFDM Based on Support Vector Machine,TN919.3
  21. Study on Face Recognition Using Scale Invariant Feature Transform and Support Vector Machine,TP391.41

CLC: > Mathematical sciences and chemical > Mathematics > Computational Mathematics > Numerical Analysis > Numerical approximation
© 2012 www.DissertationTopic.Net  Mobile