图像仿射不变方法研究
2009-8
哈尔滨工程大学出版社
申丽然，尹清波 著
140
本书研究的重点是旋转不变纹理分类和尺度不变模式识别。对旋转不变纹理分类的研究，我们提出了两种新方法，即基于Gabor小波的旋转不变纹理分类方法和基于圆形Gabor小波的旋转不变纹理分类方法。同时为了测定旋转纹理的角度，我们又提出了一种基于相位一致性的方向估计算法。对尺度不变模式识别的研究，我们提出了基于经验模态分解和梅林径向调和分解的方法。 本书可作为模式识别、图像处理及相近专业的研究生和专业人员的教材或参考书。
Chapter Ⅰ Introduction 1.1 Background 1.2 The Problem of Invariance 1.3 Scope of the Research and Contribution of this Book 1.4 The Mathematical Background 1.5 Organization of This Book 1.6 SummaryChapter Ⅱ Review of the Previous Works 2.1 Review of Rotation Invariant Texture Classification 2.2 Review of the Correlation Filters about Scale Invariant Pattern Recognition 2.3 SummaryChapter Ⅲ The Proposed Invariant Methods 3.1 Proposed Rotation Invariant Texture Classification Methods[170,171,173,174] 3.2 Proposed Orientation Estimation Method[172] 3.3 Proposed Scale Invariant Pattern Recognition Method[168,169] 3.4 SummaryChapter Ⅳ Experimental Results and Analysis 4.1 Experiments for Rotation Invariant Texture Classification 4.2 Experiments for Orientation Estimation 4.3 Experiments for Scale Invariant Pattern Recognition 4.4 SummaryChapter Ⅴ Conclusion 5.1 Rotation-invariant texture classification Using Gabor wavelet 5.2 Rotation-Invariant Texture Classification Using Circular Gabor Wavelets 5.3 Orientation estimat on from phase congruency 5.4 Scale Invariant Pattern RecofnitionReference
