Lulu Pan

Super-resolution from a single image based on local self-similarity

Super-resolution from a single image plays an important role in many areas. However, it is still a challenging work, especially in the high-resolution image’s quality and the algorithm’s efficiency. To obtain high-resolution images, a new single image super-resolution technique that extends existing learning-based super-resolution frameworks is presented in this paper. We don’t use any external example database or image pyramid to learn the missing details, and propose a single image SR method by learning local self-similarities from the original image itself. To synthesize the missing details, we design new filters which based on principles that model the super-resolution process, and use the new filters to establish the HR-LR patch pairs using the original image and its downsampled version. To obtain the SR image, we adopt a gradual magnification scheme to upscale the original image to the desired size step by step. In addition, to control the iterative error, we use the original image to guide the details added. Experimental results demonstrate that the proposed method is very flexible and give good empirical results.

Feature matching based on unsupervised manifold alignment

Feature-based methods for image registration frequently encounter the correspondence problem. In this paper, we formulate feature-based image registration as a manifold alignment problem, and present a novel matching method for finding the correspondences among different images containing the same object. Different from the semi-supervised manifold alignment, our methods map the data sets to the underlying common manifold without using correspondence information. An iterative multiplicative updating algorithm is proposed to optimize the objective, and its convergence is guaranteed theoretically. The proposed approach has been tested for matching accuracy, and robustness to outliers. Its performance on synthetic and real images is compared with the state-of-the-art reference algorithms.

Spectral feature matching based on partial least squares

We investigate the spectral approaches to the problem of point pattern matching, and present a spectral feature descriptors based on partial least square (PLS). Given keypoints of two images, we define the position similarity matrices respectively, and extract the spectral features from the matrices by PLS, which indicate geometric distribution and inner relationships of the keypoints. Then the keypoints matching is done by bipartite graph matching. The experiments on both synthetic and real-world data corroborate the robustness and invariance of the algorithm.

Feature matching using modified projective nonnegative matrix factorization

We present a novel matching method to find the correspondences among different images containing the same object. In the proposed method, by considering each feature point-set as a matrix, two point-sets are projected onto a common subspace using modified projective nonnegative matrix factorization. The core idea of the proposed approach is to jointly factorize of the two feature matrices and the matching operate on embeddings of the two point-sets in the common subspace. Furthermore, it is robust to noise due to the merit of the subspace method. The proposed approach was tested for matching accuracy, and robustness to noise. Its performance on synthetic and real images was compared with state-of-the-art reference algorithms.

Single image super resolution based on multiscale local similarity and neighbor embedding

Image quality and algorithm efficiency are the two core problems of super resolution (SR) from a single image. In this paper, we propose a novel single image SR method by using multiscale local similarity and neighbor embedding method. The proposed algorithm utilizes the self similarity redundancy in the original input image, and does not depend on external example images or the whole input image to search and match patches. Instead, we search and match patches in a localized region of the image in each level, which can improve the algorithm efficiency. The neighbor embedding method is used to generate more accurate patches for reconstruction. Finally, we use the original image and filters we design to control the iterate errors which caused by layered reconstruction, which can further improve the quality of SR results. Experimental results demonstrate that our method can ensure the quality of SR images and improve the algorithm efficiency.

Unsupervised change detection in SAR images based on frequency difference and a modified fuzzy c-means clustering

ABSTRACT Change detection for synthetic aperture radar (SAR) images is a key process in many applications exploiting remote-sensing images. It is a challenging task due to the presence of speckle noise in SAR imaging. This article investigates the problem of change detection in multitemporal SAR images. Our motivation is to avoid using only one detector to measure the change level of different features which is usually considered by classical methods. In this article, we propose an unsupervised change detection approach based on frequency difference in wavelet domain and a modified fuzzy c-means (FCM) clustering algorithm. First, the proposed method extracts high-frequency and low-frequency components using wavelet transform, and then constructs high-frequency and low-frequency difference images using different detectors. Finally, inverse wavelet transform is carried out to obtain the final difference image. In addition, inspired by manifold structure constraint, we incorporate weighted local information into the FCM to reduce the influence of speckle noise. Experimental results performed on simulated and real SAR images show the effectiveness of the proposed method, in terms of detection performance, compared with the state-of-the-art methods.

Manifold-Preserving Common Subspace Factorization for Feature Matching

A method, called Manifold-preserving Common Subspace Factorization, is presented which can be used for feature matching. Motivated by the Graph Regularized Non-negative Matrix Factorization (GNMF) algorithm (Deng et al. 2011), we developed GNMF algorithm by considering a joint factorization of the two feature matrices, which share a common basis matrix. An iterative multiplicative updating algorithm is proposed to optimize the objective, and its convergence is guaranteed theoretically. Our feature matching algorithm operates on the new representations in the common subspace generated by basis vectors. Experiments are conducted on the synthetic and real-world data. The results show that the Manifold-preserving common subspace factorization algorithm provides better matching rates than other matrix-factorization techniques.

Point pattern matching based on kernel partial least squares

Point pattern matching is an essential step in many image processing applications. This letter investigates the spectral approaches of point pattern matching, and presents a spectral feature matching algorithm based on kernel partial least squares (KPLS). Given the feature points of two images, we define position similarity matrices for the reference and sensed images, and extract the pattern vectors from the matrices using KPLS, which indicate the geometric distribution and the inner relationships of the feature points. Feature points matching are done using the bipartite graph matching method. Experiments conducted on both synthetic and real-world data demonstrate the robustness and invariance of the algorithm.

Point pattern matching based on manifold embedding

The problem of point pattern matching (PPM) is frequently encountered in computer vision, such as image registration and image matching. This paper investigates the manifold approaches to the problem of point pattern matching, and proposes a manifold correspondence based on Locally Linear Embedding (LLE). Our method operates on embeddings of the two data sets in the manifold space so as to get embedding features, which is invariance to rotation, scaling and translation (RST). By comparing the manifold embeddings of the points, we locate correspondences. We evaluate the method on both synthetic and real-world data, and experimental results demonstrate its high accuracy and robust to outliers.

A new method for texture classification and its application in texture synthesis

This paper gives an analysis of the matching process in patch-based sampling texture synthesis algorithm, and finds that the matching process takes up most of the time when synthesizing textures. But for a certain type of textures, with their texture features distributed uniformly, we can omit the matching process when they are synthesized, or make the matching roughly. We propose a novel character of images called balance coefficient that can be used to calculate texture standard coefficient and estimate texture standard property, and apply it to texture synthesis. Compared to the time complexity of general patch-based sampling algorithm, our algorithm can accelerate the synthesis and obtain a good result.