封面
声明
中文摘要
英文摘要
目录
Chapter 1 Introduction
1.1 Background
1.2 Our Research
1.3 Thesis Organization
Chapter 2 Related Work
Chapter 3 Saliency Map Improvement Based on Probability Density and Image Reconstruction Error
3.1 Process of Our Research
3.1.1 Gradient Based Saliency Map
3.1.2 Contrast Based Saliency Map Using Gaussian Image Pyramid
3.1.3 Color and DMT-based Saliency Map
3.1.5 Saliency Detection Based on Dense and Sparse Construction Error
3.1.6 Top-down Visual Saliency Detection via Machine Learning
3.2 Improve Saliency Map Using Probability Density
3.2.1 How to Improve the Saliency Map
3.2.2 Introduction of KNN Algorithm
3.2.3 Using KNN to Get the Probability Density
3.2.4 Improve Saliency Map by a Strengthen Function
3.2.5 Framework of Our Proposed Method
3.2.6 Conclusion
3.3 Improve Saliency Map Using Sparse Reconstruction Error
3.3.1 Choose Salient Templates Based on Spectral Residual
3.3.2 Calculate Sparse Reconstruction Error As the Saliency
3.3.3 Framework of Our Proposed Method
3.3.4 Conclusion
Chapter 4 Improved Seam Carving Using Random Walk Algorithm
4.1 Introduction of Seam Carving Algorithm
4.2 Our Improvement of Seam Carving Algorithm
4.2.1 Estimate the Image Energy Function
4.2.2 Random Walk Model
4.2.3 Find the Optimal Seam using Random Walk Algorithm
4.3 Grid-based Image Enlarging Based on Our Saliency Map
4.4 Experimental Results
4.4.1 Experimental Results
4.4.2 Conclusion
Chapter 5 Conclusions and Future Work
5.1 Conclusions
5.2 Future Work
参考文献
致谢
