Affine SoftAssign with bidirectional distance for point matching

Robust point cloud registration based on both hard and soft assignments

Computer vision and image recognition became one of the interesting research areas. Image registration has been widely used in fields such as computer vision, MRI images, and face recognition. Image registration is a process of aligning multiple images of the same scene which are taken from a different angle or at a different time to the same coordinate system. Image registration transforms the target image to the source image based on the affine transformation such as translation, scaling, reflection, rotation, shearing etc. It is a challenging task to find enough matching points between the source and the target images. In the proposed method, we used Speeded-Up Robust Features (SURF) and Random sample consensus (RANSAC) to find the best matching points between the pair images in addition to the minimized cost function which enhances the image registration with a few matching points. We took in our concentration some of the affine transformation which is translation, rotation, and scaling. We achieved a higher accuracy in the image registration with few matching points as low as two matching points. Experimental results show the efficiency and effectiveness of the proposed method.

To resolve multi-sensor remote sensing images registration, a new approach which combines feature and regional similarity measure is proposed. This approach firstly adopts SIFT algorithm to match feature points and constructs initial affine transformation function; next uses normalized cross-correlation coefficient (NCC) to define regional similarity measure, and improves NCC with initial affine transformation parameters in order to get further matching points from those unmatched in SIFT algorithm. Then, after gross filtering the effective matching feature points, builds the model for image registration by all the matching feature points. This method combines the advantages of feature matching and regional matching, and solves the problem that the multi-sensor remote sensing images are difficult to match correctly because of geometric and radiometric differences. Experiments show that this method has strong robustness, and are of higher registration accuracy than single registration method.

To improve the effectiveness of dose distributions in radiotherapy planning and treatment, we investigated a robust image registration algorithm based on control volumes. It auto-mapps control volumes into Thin Plate Spline deformable transformation. The control volumes are introduced to increase the accuracy for the deformable registration. We also developed a bi-directional mapping method to choose the effective control volumes, which furthermore increase the registration accuracy. The clinical images are simulated using our two-step image registration algorithm. The simulation results indicate that the registration algorithm is robust and universal, even in great deformation cases. The algorithm provides a more convenient and robust resolution as compared to manual landmark based methods. The proposed method will help towards realizing the goal of automatic image registration in radiation therapy.

Dynamic enhancement causes serious problems for registration of contrast enhanced breast MRI, due to variable uptakes of agent on different tissues or even same tissues in the breast. We present an iterative optimization algorithm to de-enhance the dynamic contrast-enhanced breast MRI and then register them for avoiding the effects of enhancement on image registration. In particular, the spatially varying enhancements are modeled by a Markov Random Field, and estimated by a locally smooth function with boundaries using a graph cut algorithm. The de-enhanced images are then registered by conventional B-spline based registration algorithm. These two steps benefit from each other and are repeated until the results converge. Experimental results show that our two-step registration algorithm performs much better than conventional mutual information based registration algorithm. Also, the effects of tumor shrinking in the conventional registration algorithms can be effectively avoided by our registration algorithm.

Three‐dimensional (3D) point cloud registration is a fundamental key issue in 3D reconstruction, 3D object recognition and augmented reality. In this study, the authors propose a novel local feature descriptor called local angle statistics histogram (LASH) for efficient 3D point cloud registration. LASH forms a description of local shape geometries by encoding their properties on angles between the normal vector of the point and the vector formed by the point and other points in its local neighbourhood. In addition, they propose a 3D point cloud registration algorithm based on LASH. The registration algorithm firstly detects triangle matching points with consistent similarity ratios, and then aggregates each pair of triangular matching points into a set of matching points. They can use these matching sets to calculate multiple transformations between two point clouds. Finally, they use the error function to identify the best transformation and to achieve coarse alignment of the two point clouds. Experiments and comparisons with other global algorithms demonstrate that the proposed approach can be applied to register point clouds with considerable or limited overlaps and is robust to noise.

Equivalency of convergence between one-step iteration algorithm and two-step iteration algorithm of variational inclusions for [formula omitted]-monotone mappings

Robust probability model based on variational Bayes for point set registration

As an extension of the classic rigid registration algorithm-Iterative Closest Point (ICP) algorithm, this paper proposes a new non-rigid ICP algorithm to match two point sets. Each point in the data set is supposed to match to the model set via an affine transformation. The proposed registration model is built up with a regularization term based on their average affine transformation. For each iteration of our algorithm, firstly correspondences between two point sets are built by the nearest-point search. Then the non-rigid transform parameters between two correspondence point sets are estimated by the proposed method in the closed form. Finally the average affine transformation is updated. A set of challenging data including single and overlapping chromosome images are tested which have significant local non-rigid transformations. Experimental results demonstrate our algorithm has higher accuracy and faster rate of convergence than other algorithms.

Image registration technology has been gradually applied in military and civil fields, such as unmanned aerial vehicle (UAV) target recognition, remote sensing image registration, 3D object reconstruction and so on. The registration accuracy of traditional registration algorithms, such as sift and surf, are not high and even mismatch, when the image content changes or affine transformation occurs. In order to solve this problem, an image registration method (IR-GAN method) based on Generative adversarial networks (GAN) was proposed, in view of the content change and affine transformation of images for training. At the same time, the sample images are expanded by translation, rotation, reflection, scaling, shearing, brightness transformation to improve the precision and accuracy of the registration algorithm. Simulation results show that our method is correct in principle and can improve the registration accuracy of both images with content changing and affine transform images.

The problem of aligning of 3D point data is the known registration task. The most popular registration algorithm is the Iterative Closest Point (ICP) algorithm. The traditional ICP algorithm is a fast and accurate approach for rigid registration between two point clouds but it is unable to handle affine case. Recently, extension of the ICP algorithm for composition of scaling, rotation, and translation is proposed. A generalized ICP version for an arbitrary affine transformation is also suggested. In this paper, a new iterative algorithm for registration of point clouds based on the point-to-plane ICP algorithm with affine transformations is proposed. At each iteration, a closed-form solution to the affine transformation is derived. This approach allows us to get a precise solution for transformations such as rotation, translation, and scaling. With the help of computer simulation, the proposed algorithm is compared with common registration algorithms.

The traditional affine iterative closest point (ICP) algorithm is fast and accurate for affine registration between two point sets, but it is easy to fall into local minimum. This paper proposes a robust Affine ICP algorithm based on corner points. First, an objective function is established under the guidance of corner points, where the corner points as the shape control point guides the affine registration of 2D point sets. Then, at each step of the algorithm, the affine transformation obtained by the last iterative step is used to establish the correspondence of these two point sets. Next, the new affine transformation is solved by using the objective function under the guidance of the corner points. Experimental results demonstrate that the robustness and convergence of our algorithm are greatly improved compared with the traditional affine registration algorithm.

In allusion to non-rigid registration of medical images, the paper gives a novel algorithm based on improved Scale Invariant Features Transform (SIFT) feature matching algorithm. First, Harris corner detection algorithm is used in the process of scale invariant feature extraction, so the number of right matching points is increased; with regard to the feature points detected in the scale space, an improved SIFT feature extraction algorithm with global context vector is presented to solve the problem that SIFT descriptors result in a lot of mismatches when an image has many similar regions. On this basis, affine transformation is chosen to implement the non-rigid registration, and weighted mutual information (WMI) measure and Particle Swarm Optimization (PSO) algorithm are also chosen to optimize the registration process. The experimental results show that the method can achieve better registration results than the method based on mutual information.

Point set registration is a key problem in intelligent vehicle localization and mapping. This paper presents a rigid point set registration algorithm based on the cubature Kalman filter (CKF). First, the point set registration problem is cast into the state space model assuming that the correspondence between these two point sets is previously unknown. Then, CKF is used to solve this nonlinear filtering problem. At every iterative step, all the points in the moving point set will be considered, and the corresponding points in the model point set will be updated accordingly. In the time update, the scale of the free space that can be explored is significant for this registration algorithm. Herein, continuous simulated annealing (CSA) is adopted to gradually optimize the covariance of the model noise to speed up convergence. Tests on public data sets show that the CKF-based point set registration algorithm is robust to outliers, noise, and initialization misalignment. Then, the application of this algorithm on intelligent vehicles and intelligent transportation systems is demonstrated through mapping and localization experiments. The precision and robustness are both validated compared with the traditional ICP, NDT, and CPD based point set registrations in the localization experiments. Thus, the contributions of this paper are threefold. First, a CKF scheme is utilized in the point set registration problem. Second, CSA serves as the local optimizer to speed up the convergence process and make it more accurate. Third, this filtering-based method is adopted in intelligent vehicles and SLAM applications.

Genetic Algorithms (GAs) have been applied to matching problem. However, traditional GAs do not perform well in matching problem because there can be many locally similar parts. This paper presents a new genetic algorithm for silhouette matching. New concepts of partially matched genestrings in the initial population, the extending operator and the order adjustment algorithm are proposed. Each gene-string in the initial population only has three matched points while other points are unmatched. During the evolution, each gene-string will have more matched points due to the applications of the crossover and extending operators. The extending operator determines a potential match for an unmatched point near a matched point by searching the local space. After the application of the crossover and extending operators, the adjustment algorithm enforces each gene-string to be an ordered list by removing some matched points, if necessary. Our experiments show that the new matching algorithm based on GA performs better than traditional GA-based algorithms.