Feature-based Registration Techniques Research Articles

Imaging and Laser Scanning–Based Noncontact Deflection Monitoring Technique for Timber Railroad Bridges

This study developed a noncontact deflection monitoring technique using terrestrial laser scanner (TLS) and images for faster, safer, and efficient 2D/3D bridge deflection measurements. The technique was applied to assess the serviceability performance of two timber trestle railroad bridges located in Florida and validated based on bridge deflections from deflectometers. A key challenge includes extracting control points that are visible both in images and TLS data due to different modalities of the data. Hence, this study presents a method of extracting linear features from both images and TLS data. The camera pose was derived from images and TLS data by using a linear feature-based registration algorithm. The deformations in the structure were then detected by measuring the points of interest for different loads. There are several unique contributions of the study. First, there is no requirement for the use of targets which improves the safety of bridge engineers. Second, it is a relatively cost-effective technique for obtaining bridge deflections due to moving train loads, as the use of a laser scanner as part of the noncontact technique is only once in a lifetime for a given bridge. Alternatively, total stations can also be used to capture linear features. Finally, this technique produced accurate deflection measurements using a linear feature-based registration technique. This study found that the noncontact deflection monitoring technique agrees well with actual deformations observed from deflectometers.

Linear Feature-Based Image/LiDAR Integration for a Stockpile Monitoring and Reporting Technology

Stockpile monitoring has been recently conducted with the help of modern remote sensing techniques – e.g., terrestrial/aerial photogrammetry/LiDAR – that can efficiently produce accurate 3D models for the area of interest. However, monitoring of indoor stockpiles still requires more investigation due to unfavorable conditions in these environments such as lack of global navigation satellite system (GNSS) signals and/or homogenous texture. This study develops a fully-automated image/LiDAR integration framework that is capable of generating accurate 3D models with color information for stockpiles under challenging environmental conditions. The derived colorized 3D point cloud can be subsequently used for volume estimation and visual inspection of stockpiles. The main contribution of the developed strategy is using automatically derived conjugate image/LiDAR linear features for simultaneous registration and camera/LiDAR system calibration. Data for this study is acquired using a camera-assisted LiDAR mapping platform – denoted as stockpile monitoring and reporting technology (SMART) – which was recently designed as a time-efficient and cost-effective bulk material tracking. Experimental results on three datasets show that the developed framework outperforms a classical planar feature-based registration technique in terms of the alignment of acquired point cloud. Results also indicate that the proposed approach can lead to a high relative accuracy between image lines and their corresponding back-projected LiDAR features in the range of 4–7 pixels.

Scan registration using planar features

Abstract. Point cloud acquisition by using laser scanners provides an efficient way for 3D as-built modelling of indoor/outdoor urban environments. In the case of large structures, multiple scans may be required to cover the entire scene and registration is needed to merge them together. In general, the identification of corresponding geometric features among a series of scans can be used to compute the 3D rigid-body transformation useful for the registration of each scan into the reference system of the final point cloud. Different automatic or semi-automatic methods have been developed to this purpose. Several solutions based on artificial targets are available, which however may not be suitable in any situations. Methods based on surface matching (like ICP and LS3D) can be applied if the scans to align have a proper geometry and surface texture. In the case of urban and architectural scenes that present the prevalence of a few basic geometric shapes ("Legoland" scenes) the availability of many planar features is exploited here for registration. The presented technique does not require artificial targets to be added to the scanned scene. In addition, unlike other surface-based techniques (like ICP) the planar feature-based registration technique is not limited to work in a pairwise manner but it can handle the simultaneous alignment of multiple scans. Finally, some applications are presented and discussed to show how this technique can achieve accuracy comparable to a consolidated registration method.

Lung registration using automatically detected landmarks.

Accurate registration of lung CT images is inevitable for numerous clinical applications. Usually, nonlinear intensity-based methods are used. Their accuracy is typically evaluated using corresponding anatomical points (landmarks; e.g. bifurcations of bronchial and vessel trees) annotated by medical experts in the images to register. As image registration can be interpreted as correspondence finding problem, these corresponding landmarks can also be used in feature-based registration techniques. Recently, approaches for automated identification of such landmark correspondences in lung CT images have been presented. In this work, a novel combination of variational nonlinear intensity-based registration with an approach for automated landmark correspondence detection in lung CT pairs is presented and evaluated. The main blocks of the proposed hybrid intensity- and feature-based registration scheme are a two-step landmark correspondence detection and the so-called CoLD (Combining Landmarks and Distance Measures) framework. The landmark correspondence identification starts with feature detection in one image followed by a blockmatching-based transfer of the features to the other image. The established correspondences are used to compute a thin-plate spline (TPS) transformation. Within CoLD, the TPS transformation is improved by minimization of an objective function consisting of a Normalized Gradient Field distance measure and a curvature regularizer; the landmark correspondences are guaranteed to be preserved by optimization on the kernel of the discretized landmark constraints. Based on ten publicly available end-inspiration/expiration CT scan pairs with anatomical landmark sets annotated by medical experts from the DIR-Lab database, it is shown that the hybrid registration approach is superior in terms of accuracy: The mean distance of expert landmarks is decreased from 8.46 mm before to 1.15 mm after registration, outperforming both the TPS transformation (1.68 mm) and a nonlinear registration without usage of automatically detected landmarks (2.44 mm). The improvement is statistically significant in eight of ten datasets in comparison to TPS and in nine of ten datasets in comparison to the intensity-based registration. Furthermore, CoLD globally estimates the breathing-induced lung volume change well and results in smooth and physiologically plausible motion fields of the lungs. We demonstrated that our novel landmark-based registration pipeline outperforms both TPS and the underlying nonlinear intensity-based registration without landmark usage. This highlights the potential of automatic landmark correspondence detection for improvement of lung CT registration accuracy.

Geodesics-Based Image Registration: Applications To Biological And Medical Images Depicting Concentric Ring Patterns

In many biological or medical applications, images that contain sequences of shapes are common. The existence of high inter-individual variability makes their interpretation complex. In this paper, we address the computer-assisted interpretation of such images and we investigate how we can remove or reduce these image variabilities. The proposed approach relies on the development of an efficient image registration technique. We first show the inadequacy of state-of-the-art intensity-based and feature-based registration techniques for the considered image datasets. Then, we propose a robust variational method which benefits from the geometrical information present in this type of images. In the proposed non-rigid geodesics-based registration, the successive shapes are represented by a level-set representation, which we rely on to carry out the registration. The successive level sets are regarded as elements in a shape space and the corresponding matching is that of the optimal geodesic path. The proposed registration scheme is tested on synthetic and real images. The comparison against results of state-of-the-art methods proves the relevance of the proposed method for this type of images.

Experimental validation of an intrasubject elastic registration algorithm for dynamic‐3D ultrasound images

In image-guided therapy, real-time visualization of the anatomy and adjustments in the therapy plan due to anatomical motions during the procedure is of outmost importance. 3D ultrasound has the potential to enable this real-time monitoring; however, nonrigid registration of a sequence of 3D ultrasound volumes remains to be a challenging problem. The authors present our recent results on the development of a computationally inexpensive feature-based registration algorithm for elastic alignment of dynamic-3D ultrasound images. Our algorithm uses attribute vectors, based on the image intensity and gradient information, to perform feature-based matching in a sequence of 3D ultrasound images. Prior information from both the fixed and previous moving images is utilized to track features throughout the 3D image series. The algorithm has been compared to various publicly available registration techniques, i.e., the B-splines deformable registration, the symmetric forces Demons, and the fast free-form deformable registration method. Using a series of validation experiments on datasets collected from carotid artery, liver, and kidney of 20 subjects, the authors demonstrate that the feature-based, B-splines, Demons, and fast free-form deformable registration techniques can all recover volume deformations in a 3D ultrasound image series with reasonable accuracy; however, the proposed feature-based registration technique has substantial computational advantage over the other approaches. The proposed feature-based registration technique has the potential for real-time implementation on a computationally inexpensive platform and has the capability of recovering nonrigid deformations in tissue with reasonable accuracy.

조영 전후의 폐 CT 영상 정합을 위한 특징 기반의 비강체 정합 기법

In this paper, a feature-based registration technique is proposed for pre-contrast and post-contrast lung CT images. It utilizes three dimensional(3-D) features with their descriptors and estimates feature correspondences by nearest neighborhood matching in the feature space. We design a transformation model between the input image pairs using a free form deformation(FFD) which is based on B-splines. Registration is achieved by minimizing an energy function incorporating the smoothness of FFD and the correspondence information through a non-linear gradient conjugate method. To deal with outliers in feature matching, our energy model integrates a robust estimator which discards outliers effectively by iteratively reducing a radius of confidence in the minimization process. Performance evaluation was carried out in terms of accuracy and efficiency using seven pairs of lung CT images of clinical practice. For a quantitative assessment, a radiologist specialized in thorax manually placed landmarks on each CT image pair. In comparative evaluation to a conventional feature-based registration method, our algorithm showed improved performances in both accuracy and efficiency.

Real-Time Multimodal Retinal Image Registration for a Computer-Assisted Laser Photocoagulation System

An algorithm for the real-time registration of a retinal video sequence captured with a scanning digital ophthalmoscope (SDO) to a retinal composite image is presented. This method is designed for a computer-assisted retinal laser photocoagulation system to compensate for retinal motion and hence enhance the accuracy, speed, and patient safety of retinal laser treatments. The procedure combines intensity and feature-based registration techniques. For the registration of an individual frame, the translational frame-to-frame motion between preceding and current frame is detected by normalized cross correlation. Next, vessel points on the current video frame are identified and an initial transformation estimate is constructed from the calculated translation vector and the quadratic registration matrix of the previous frame. The vessel points are then iteratively matched to the segmented vessel centerline of the composite image to refine the initial transformation and register the video frame to the composite image. Criteria for image quality and algorithm convergence are introduced, which assess the exclusion of single frames from the registration process and enable a loss of tracking signal if necessary. The algorithm was successfully applied to ten different video sequences recorded from patients. It revealed an average accuracy of 2.47 ± 2.0 pixels (∼23.2 ± 18.8 μm) for 2764 evaluated video frames and demonstrated that it meets the clinical requirements.

Hand-held 3D scanning based on coarse and fine registration of multiple range images

A hand-held 3D scanning technique is proposed to reconstruct 3D models of real objects. A sequence of range images captured from a hand-held stereo camera is automatically registered to a reference coordinate system. The automated scanning process consists of two states, coarse and fine registration. At the beginning, scanning process starts at the fine registration state. A fast and accurate registration refinement technique is used to align range images in a pair-wise manner. If the refinement technique fails, the process changes to the coarse registration state. A feature-based coarse registration technique is proposed to find correspondences between the last successful frame and the current frame. If the coarse registration successes, the process returns to the fine registration state again. A fast point-to-plane refinement technique is employed to do shape-based registration. After the shape-based alignment, a texture-based refinement technique matches texture features to enhance visual appearance of the reconstructed models. Through a graphic and video display, a human operator adjusts the pose of the camera to change the view of the next acquisition. Experimental results show that 3D models of real objects are reconstructed from sequences of range images.

Inside the beating heart: an in vivo feasibility study on fusing pre- and intra-operative imaging for minimally invasive therapy

An interventional system for minimally invasive cardiac surgery was developed for therapy delivery inside the beating heart, in absence of direct vision. A system was developed to provide a virtual reality (VR) environment that integrates pre-operative imaging, real-time intra-operative guidance using 2D trans-esophageal ultrasound, and models of the surgical tools tracked using a magnetic tracking system. Detailed 3D dynamic cardiac models were synthesized from high-resolution pre-operative MR data and registered within the intra-operative imaging environment. The feature-based registration technique was employed to fuse pre- and intra-operative data during in vivo intracardiac procedures on porcine subjects. This method was found to be suitable for in vivo applications as it relies on easily identifiable landmarks, and hence, it ensures satisfactory alignment of pre- and intra-operative anatomy in the region of interest (4.8 mm RMS alignment accuracy) within the VR environment. Our initial experience in translating this work to guide intracardiac interventions, such as mitral valve implantation and atrial septal defect repair demonstrated feasibility of the methods. Surgical guidance in the absence of direct vision and with no exposure to ionizing radiation was achieved, so our virtual environment constitutes a feasible candidate for performing various off-pump intracardiac interventions.

Robust feature-based registration using a Gaussian-weighted distance map and brain feature points for brain PET/CT images

Feature-based registration is an effective technique for clinical use, because it can greatly reduce computational costs. However, this technique, which estimates the transformation by using feature points extracted from two images may cause misalignments, particularly in brain PET and CT images that have low correspondence rates between features due to differences in image characteristics. To cope with this limitation, we propose a robust feature-based registration technique using a Gaussian-weighted distance map (GWDM) that finds the best alignment of feature points even when features of two images are mismatched. A GWDM is generated by propagating the value of the Gaussian-weighted mask from feature points of CT images and leads the feature points of PET images to be aligned on an optimal location even though there is a localization error between feature points extracted from PET and CT images. Feature points are extracted from two images by our automatic brain segmentation method. In our experiments, simulated and clinical data sets were used to compare our method with conventional methods such as normalized mutual information (NMI)-based registration and chamfer matching in accuracy, robustness, and computational time. Experimental results showed that our method aligned the images robustly even in cases where conventional methods failed to find optimal locations. In addition, the accuracy of our method was comparable to that of the NMI-based registration method.

Surface parameterization in volumetric images for curvature-based feature classification

Curvature-based surface features are well suited for use in multimodal medical image registration. The accuracy of such feature-based registration techniques is dependent upon the reliability of the feature computation. The computation of curvature features requires second derivative information that is best obtained from a parametric surface representation. We present a method of explicitly parameterizing surfaces from volumetric data. Surfaces are extracted, without a global thresholding, using active contour models. A monge/spl acute/ basis for each surface patch is estimated and used to transform the patch into local, or parametric, coordinates. Surface patches are fit to a bicubic polynomial in local coordinates using least squares solved by singular value decomposition. We tested our method by reconstructing surfaces from the surface model and analytically computing Gaussian and mean curvatures. The model was tested on analytical and medical data.