Bobkov A.V. - Image registration in the real time applications
Описание файла
PDF-файл из архива "Bobkov A.V. - Image registration in the real time applications", который расположен в категории "". Всё это находится в предмете "распознавание изображений" из 10 семестр (2 семестр магистратуры), которые можно найти в файловом архиве МГТУ им. Н.Э.Баумана. Не смотря на прямую связь этого архива с МГТУ им. Н.Э.Баумана, его также можно найти и в других разделах. Архив можно найти в разделе "книги и методические указания", в предмете "распознавание изображений" в общих файлах.
Просмотр PDF-файла онлайн
Текст из PDF
De Montfort UniversityLeicesterFEATURE-BASEDIMAGE REGISTRATIONIN REAL TIME APPLICATIONSSeptember 2007PhD ThesisAlexander Bobkov2007ContentsCHAPTER 1. INTRODUCTION AND LITERATURE REVIEW..................101.1. INTRODUCTION.............................................................................. 101.1.1. Specifics of aerial image registration.................................
111.1.2. A brief review of image analysis tasks...............................121.1.3. Aims and objectives........................................................... 151.1.4. Thesis outline..................................................................... 161.2. IMAGE REGISTRATION PROBLEM........................................................171.2.1. Image registration: statement of the problem.....................181.2.2. Classification of image registration methods..................... 191.2.2.1.
Space transformation................................................... 191.2.2.1.1 Global transformation............................................191.2.2.1.2. Local transformation.............................................201.2.2.1.3. Optical flow...........................................................201.2.2.2. Image representation ................................................... 211.2.2.2.1. Image representation methods – an overview.......221.2.2.2.2. Boundary representation.......................................231.2.2.2.3. Region-based representation................................
251.2.2.2.4. Summary on image representation........................271.2.2.3. Optimisation strategy................................................... 281.2.2.3.1. Optimality criterion...............................................281.2.2.3.2. Optimisation algorithm......................................... 291.2.2.3.3. Exhaustive search..................................................301.2.2.3.4.
Graph-oriented search.......................................... 301.2.2.3.5. Hough Transform strategy.................................... 311.3. LINE AND CURVE EXTRACTION......................................................... 321.3.1. Parameterisation scheme selection..................................... 331.3.2.
Reducing the parameter space............................................ 341.3.3. Parameter space decomposition......................................... 351.3.4. Using additional information..............................................351.3.5. Using different analysing orders ....................................... 351.3.6. Line splitting problem........................................................ 361.3.7. Line segment detection.......................................................371.4.
CONTOUR DETECTION..................................................................... 381.4.1. Edge detector...................................................................... 391.4.2. Edge labelling.....................................................................421.4.2.1. Edge thinning............................................................... 421.4.2.1.1. Skeletonization...................................................... 431.4.2.1.2. Non-maximum suppression................................... 431.4.2.2. Thresholding................................................................ 431.5. SUMMARY ...................................................................................
44CHAPTER 2. IMAGE CONTOUR EXTRACTION.........................................4522.1. DIFFERENTIAL METHOD...................................................................472.2. STATISTICAL METHODS................................................................... 482.2.1. How to select the area size for a statistics collecting......... 492.2.2. How to determine a threshold and amount of modes......... 492.2.3. Statistical method characteristics....................................... 502.3. CONTOUR THINNING.......................................................................
502.3.1. Thresholding.......................................................................522.3.2. Global and local thresholding comparison......................... 542.3.3. Non-maximum suppression................................................572.3.4. Use the thresholding and edge thinning at the same time.. 592.4. COMPARISON OF STATISTICAL AND DIFFERENTIAL METHODS..................
592.5. CONCLUSION.................................................................................65CHAPTER 3. LINE AND CORNER DETECTION..........................................673.1. CONTOUR REPRESENTATION ............................................................ 673.2. LINE DETECTION............................................................................683.3. ARC AND CURVE DETECTION............................................................693.4. CORNER DETECTION....................................................................... 703.5. LINE SEGMENT DETECTION..............................................................
723.6. PROPOSED METHOD OF LINE SEGMENT DETECTION............................... 743.6.1. Extended Hough transformation........................................ 743.6.2. Algorithm based on Extended Hough Transform.............. 763.6.3. Performance of proposed algorithm................................... 783.6.4. Gluing of line segments......................................................783.7. EXPERIMENTS............................................................................... 803.7.1. Performance analysis..........................................................813.7.2. A metrics problem..............................................................
833.7.3. An optimal discretization step selection.............................853.7.4. Minimal object size that can be approximated by line segments883.8. CONCLUSION.................................................................................88CHAPTER 4. IMAGE REGISTRATION.......................................................... 904.1. INTRODUCTION.............................................................................. 904.2.
STATEMENT OF THE PROBLEM.......................................................... 904.3. IMAGE REGISTRATION STRATEGIES.................................................... 914.3.1. Exhaustive search .............................................................. 914.3.2. Graph matching ................................................................. 914.3.3.
Hough Transform strategy..................................................924.4. HOUGH TRANSFORM STRATEGY VARIATIONS...................................... 934.4.1. Point features formed from non-point ones .......................934.4.2. The search space decomposition ....................................... 944.4.3. Standard Hough Transformation approach........................
944.5. EXHAUSTIVE SEARCH......................................................................954.6. METHODS OF ENDLESS LINE COMPARISON................. 954.6.1. Line matching based on Hough transform......................... 9534.6.2. Method based on line pair comparison...............................974.7. SECONDARY FEATURE EXTRACTION METHODS........ 994.7.1. Corner comparison............................................................. 994.7.2. Stripe comparison ............................................................1004.7.3. Comparison of line segment middles...............................
1014.8. METHODS BASED ON LINE SEGMENT COMPARISON1024.8.1. Image registration by line segment pairs..........................1024.8.1.1. Looking for a line pairs..............................................1034.8.1.2. Matching pairs........................................................... 1034.8.1.3. The translation value computation.............................1044.8.1.4. Method characteristics............................................... 1064.8.2. Image registration by line segments................................. 1074.8.2.1.
Determination of transformation parameters............. 1074.8.2.4. Method characteristics............................................... 1104.9. METHOD COMPARISON..................................................................1104.9.1. Performance......................................................................1114.9.2. Accuracy...........................................................................1154.9.3. Reliability.........................................................................
1164.10. CONCLUSION.............................................................................117CHAPTER 5. DISCUSSION.............................................................................. 1185.1. CHARACTERISTICS OF IMAGE REGISTRATION METHODS........................1185.1.1.