Iterative Closest Point Method Research Articles

Locally rigid, vessel-based registration for laparoscopic liver surgery.

PurposeLaparoscopic liver resection has significant advantages over open surgery due to less patient trauma and faster recovery times, yet is difficult for most lesions due to the restricted field of view and lack of haptic feedback. Image guidance provides a potential solution but is challenging in a soft deforming organ such as the liver. In this paper, we therefore propose a laparoscopic ultrasound (LUS) image guidance system and study the feasibility of a locally rigid registration for laparoscopic liver surgery.MethodsWe developed a real-time segmentation method to extract vessel centre points from calibrated, freehand, electromagnetically tracked, 2D LUS images. Using landmark-based initial registration and an optional iterative closest point (ICP) point-to-line registration, a vessel centre-line model extracted from preoperative computed tomography (CT) is registered to the ultrasound data during surgery.ResultsUsing the locally rigid ICP method, the RMS residual error when registering to a phantom was 0.7 mm, and the mean target registration error (TRE) for two in vivo porcine studies was 3.58 and 2.99 mm, respectively. Using the locally rigid landmark-based registration method gave a mean TRE of 4.23 mm using vessel centre lines derived from CT scans taken with pneumoperitoneum and 6.57 mm without pneumoperitoneum.ConclusionIn this paper we propose a practical image-guided surgery system based on locally rigid registration of a CT-derived model to vascular structures located with LUS. In a physical phantom and during porcine laparoscopic liver resection, we demonstrate accuracy of target location commensurate with surgical requirements. We conclude that locally rigid registration could be sufficient for practically useful image guidance in the near future.

Efficient Sparse ICP

The registration of two geometric surfaces is typically addressed using variants of the Iterative Closest Point (ICP) algorithm. The Sparse ICP method formulates the problem using sparsity-inducing norms, significantly improving the resilience of the registration process to large amounts of noise and outliers, but introduces a significant performance degradation. In this paper we first identify the reasons for this performance degradation and propose a hybrid optimization system that combines a Simulated Annealing search along with the standard Sparse ICP, in order to solve the underlying optimization problem more efficiently. We also provide several insights on how to further improve the overall efficiency by using a combination of approximate distance queries, parallel execution and uniform subsampling. The resulting method provides cumulative performance gain of more than one order of magnitude, as demonstrated through the registration of partially overlapping scans with various degrees of noise and outliers.

Face recognition using a permutation coding neural classifier

Face recognition is an important security task. We propose a high-level method to solve this problem: a permutation coding neural classifier (PCNC). A PCNC with a special feature extractor for face image recognition systems is a relatively new method that has been tested with good results to classify real environment images (such as larvae of various types and handmade elements). As baseline methods, a support vector machine (SVM) and the iterative closest point (ICP) method are selected for comparison. We applied these methods to gray-level images from the FRAV3D face database. Fifteen experiments were performed to examine a large set of training and testing conditions. As a general result, it was observed that errors are lower with the PCNC than with the SVM and the ICP classifier. We aggregated various distortions for the initial images to improve the PCNC. We analyze and discuss the obtained results.

A New Method for Registration of 3D Point Sets with Low Overlapping Ratios

The registration of multi-view point sets is often used in surface reconstruction for complex parts, and many researchers are striving to improve its accuracy and efficiency. A popular algorithm for registration of 3D point sets is the Iterative Closest Point (ICP). With the increase of overlapping ratios, the convergence accuracy and speed of ICP algorithm improves. In most cases of 3D scanning systems, the overlapping regions of multi-view point sets often locate at the boundaries. Based on this characteristic, a new method for registration of multi-view point sets with low overlapping ratios is proposed. Firstly, two kinds of sampling strategies are introduced, and a great number of point pairs are constructed. Through surface classification, curvature and neighborhood curvature matching of point pairs, the overlapping regions are selected. Then a Levenberg-Marquardt ICP (LM-ICP) method is used to search optimal registration parameters with the points in the overlapping regions. Finally, experiments for registration of multi-view point sets are performed, and results show that the method can get good registration accuracy for point sets with low overlapping ratios.

Preprocedural magnetic resonance imaging for image-guided catheter ablation of scar-related ventricular tachycardia.

To present and validate a highly automated MRI analysis workflow for image-guided catheter ablation of scar-related ventricular tachycardia (VT) ablation procedures. A cohort of 15 post-infarction patients underwent MRI prior to VT ablation. The MRI study included a black-blood turbo spin echo sequence for visualizing the aortic root and ostium of the left main (LM) coronary artery, and a 3D late gadolinium enhanced sequence for visualizing the LV anatomy and myocardial scar substrate. Semi-automated segmentation of the LV, aortic root and ostium of LM was performed, followed by fully automated segmentation of myocardial scar. All segmented structures were aligned using an automated image registration algorithm to remove inter-scan displacement. MRI was integrated at the beginning of the procedure after mapping a single LM point. The integration performance was compared to that of the traditional iterative closest point (ICP) method. The proposed method required a single LM mapping point only, compared to 255 ± 43 points with the ICP method. The single-point method achieved a mean point-to-surface distance of 4.9 ± 1.5 mm on the LV surface and 5.1 ± 1.7 mm on the aorta surface (ICP: 3.7 ± 0.8 and 9.2 ± 7.2 mm, P < 0.05). The Cohen's kappa coefficient between the MRI-defined and EAM-defined scar was 0.36 ± 0.16 for the presented method, significantly higher than that of ICP method (0.23 ± 0.21, P = 0.03), indicating more accurate scar substrate localization during integration. This study demonstrated the feasibility of preprocedural MRI integration into the VT ablation procedure, with highly automated image analysis workflow and minimal mapping effort.

An enhanced method for registration of dental surfaces partially scanned by a 3D dental laser scanning

In this paper, we propose the fast and accurate registration method of partially scanned dental surfaces in a 3D dental laser scanning. To overcome the multiple point correspondence problems of conventional surface registration methods, we propose the novel depth map-based registration method to register 3D surface models. First, we convert a partially scanned 3D dental surface into a 2D image by generating the 2D depth map image of the surface model by applying a 3D rigid transformation into this model. Then, the image-based registration method using 2D depth map images accurately estimates the initial transformation between two consequently acquired surface models. To further increase the computational efficiency, we decompose the 3D rigid transformation into out-of-plane (i.e. x-, y-rotation, and z-translation) and in-plane (i.e. x-, y-translation, and z-rotation) transformations. For the in-plane transformation, we accelerate the transformation process by transforming the 2D depth map image instead of transforming the 3D surface model. For the more accurate registration of 3D surface models, we enhance iterative closest point (ICP) method for the subsequent fine registration. Our initial depth map-based registration well aligns each surface model. Therefore, our subsequent ICP method can accurately register two surface models since it is highly probable that the closest point pairs are the exact corresponding point pairs. The experimental results demonstrated that our method accurately registered partially scanned dental surfaces. Regarding the computational performance, our method delivered about 1.5 times faster registration than the conventional method. Our method can be successfully applied to the accurate reconstruction of 3D dental objects for orthodontic and prosthodontic treatment.

A new method for aspherical surface fitting with large-volume datasets

In the framework of form characterization of aspherical surfaces, European National Metrology Institutes (NMIs) have been developing ultra-high precision machines having the ability to measure aspherical lenses with an uncertainty of few tens of nanometers. The fitting of the acquired aspherical datasets onto their corresponding theoretical model should be achieved at the same level of precision. In this article, three fitting algorithms are investigated: the Limited memory-Broyden-Fletcher-Goldfarb-Shanno (L-BFGS), the Levenberg–Marquardt (LM) and one variant of the Iterative Closest Point (ICP). They are assessed based on their capacities to converge relatively fast to achieve a nanometric level of accuracy, to manage a large volume of data and to be robust to the position of the data with respect to the model. Nevertheless, the algorithms are first evaluated on simulated datasets and their performances are studied. The comparison of these algorithms is extended on measured datasets of an aspherical lens. The results validate the newly used method for the fitting of aspherical surfaces and reveal that it is well adapted, faster and less complex than the LM or ICP methods.

A hybrid point cloud alignment method combining particle swarm optimization and iterative closest point method

3D quality inspection is widely applied in many industrial fields including mould design, automotive and blade manufacturing, etc. A commonly used method is to obtain the point cloud of the inspected object and make a comparison between the point cloud and the corresponding CAD model or template. Thus, it is important to align the point cloud with the template first and foremost. Moreover, for the purpose of automatization of quality inspection, this alignment process is expected to be completed without manual interference. In this paper, we propose to combine the particle swarm optimization (PSO) with iterative closest point (ICP) algorithm to achieve the automated point cloud alignment. The combination of the two algorithms can achieve a balance between the alignment speed and accuracy, and avoid the local optimal caused by bad initial position of the point cloud.

Simulation of Registration Accuracy of Iterative Closest Point (ICP) Method for Pose Estimation

ICP method is a powerful tool widely used in 3D point registration and pose estimation application. But there lacks an deterministic relation between the 3D point dataset and the model. Here we present some Monte Carlo simulation results for pose estimation error of one degree and the system requirement for lidar 3D point measurement techniques.

Local Shape Similarity and Mean-Shift Curvature for Deformable Surface Mapping of Anatomical Structures

This paper reports a novel method for deformable registration of digital anatomical surfaces. The method capitalizes upon the iterative local affine iterative closest point (ICP) approach that applies an affine transformation per surface vertex along with a regularization constraint to force neighboring surface vertices to undergo similar transformations. More robust vertex correspondence with respect to simple closest point was obtained by exploiting local shape similarity metrics, which includes vertex distance, surface normal, and local curvature. The local curvature was mean shifted at run-time, during the iterative optimization, to make the point correspondence process less dependent upon the surface noise and resolution. The experimental validation was performed on three surface datasets (femur, hemi-pelvic bone, and liver). The registration results showed that the proposed method outperforms, across all the three surface datasets (rmse: 0.19 mm, 0.30 mm, 0.61 mm), global affine ICP (rmse: 2.89 mm, 3.95 mm, and 8.30 mm), local affine ICP (rmse: 0.31 mm, 1.61 mm, and 1.63 mm) and coherent point drift (rmse: 1.99 mm, 2.39 mm, and 4.78 mm) methods. As a whole, the mean-shifted curvature increased the registration accuracy by about 20%.

3D face alignment and registration in the presence of facial expression differences

AbstractThis paper deals with the problem of 3D alignment of faces in the presence of some facial expression changes. The data are three dimensional and obtained using a laser scanner. Our approach is based on the differential geometry of the surface and computes the intrinsic local fiducial points on the surface and on curves that reside on the surface. Because these fiducial points are local, they allow partial alignment, where part of the face is viewed. Moreover, since the fiducial points are relatively affine‐invariant to local affine transformations, they allow matching and alignment when facial expression changes affect part of the face. A fast, noniterative alignment procedure is presented in this paper that establishes reliable correspondences between fiducial points without any prior knowledge of the overall nonlinear global transformation that takes place after the changes in facial expressions. This is achieved through the construction of a set of ordered novel absolute local affine invariants. With enough fiducial points set as correspondents, the overall nonlinear transformation is computed and the face before and after the transformation are aligned. For comparison, we also compare the alignment performance of our method with that of the iterative closest point (ICP) method and the coherent point drift (CPD) method which is based on the Gaussian mixture model by looking at the between‐to‐within variation (signal‐to‐noise ratio, SNR) between these two classes (match vs nonmatch) based on the average alignment errors for the genuine matches and nonmatches, normalized by their respective within variation, and by running it on the 3D face database GavabDB. The separation between the true match and nonmatch is best for our zero‐torsion method (SNR of 1.2), followed by the ICP method (SNR of 0.37)and the CPD method (SNR of 0.15). It is interesting to observe here that, although the alignment error for the true match cases is lowest for the CPD method, it is also extremely low for the nonmatch cases, which would mean that the method would force a query into a wrong face, yielding poor specificity (lots of false alarms). © 2013 Institute of Electrical Engineers of Japan. Published by John Wiley &amp; Sons, Inc.

Fabrication and characterization of areal roughness specimens for applications in scanning probe microscopy

Reference structures are urgently required to characterize the performance of scanning probe microscopes for the quantitative measurement of micro-/nano-scale roughness. In this paper, preliminary areal roughness specimens are fabricated by electron beam lithography (EBL) and direct laser writing (DLW). The geometric topography and statistical quantities of the surface structures are specially designed. Precise registration between the design template and the measured data has been carried out to evaluate the fabrication quality. Two matching algorithms, the iterative closest point method and the Levenberg–Marquardt method, are compared. Results imply that the latter method prevails against the former one with improved accuracy and efficiency. It is demonstrated that the typical root-mean-square residuals of the EBL and DLW fabricated surfaces are approximately 43.5 and 195.3 nm, respectively. The spectral properties of the fabricated structures are also analyzed. These areal roughness specimens have a potential for validating scanning probe microscopes and other nano-measurement instruments for areal roughness measurements.

Comparing ICP variants on real-world data sets

Many modern sensors used for mapping produce 3D point clouds, which are typically registered together using the iterative closest point (ICP) algorithm. Because ICP has many variants whose performances depend on the environment and the sensor, hundreds of variations have been published. However, no comparison frameworks are available, leading to an arduous selection of an appropriate variant for particular experimental conditions. The first contribution of this paper consists of a protocol that allows for a comparison between ICP variants, taking into account a broad range of inputs. The second contribution is an open-source ICP library, which is fast enough to be usable in multiple real-world applications, while being modular enough to ease comparison of multiple solutions. This paper presents two examples of these field applications. The last contribution is the comparison of two baseline ICP variants using data sets that cover a rich variety of environments. Besides demonstrating the need for improved ICP methods for natural, unstructured and information-deprived environments, these baseline variants also provide a solid basis to which novel solutions could be compared. The combination of our protocol, software, and baseline results demonstrate convincingly how open-source software can push forward the research in mapping and navigation.

Iterative Closest Point Method with a Least Trimmed Squares Estimator

The iterative closest point (ICP) method is one of the most important methods for 2D/3D point registration. Robust statistical method is applied widely for improving the robustness of ICP. A new method that incorporates the Least Trimmed Squares (LTS) Estimator into the ICP is proposed in this paper. In this method, outliers are removed according to characteristics of residual distribution. A large number of experimental results show that the proposed method is robust and efficient.

Biomechanical Evaluation of the X-Stop Device for Surgical Treatment of Lumbar Spinal Stenosis

Controlled experimental study. To evaluate the kinematical effects of X-Stop device on the spinal process at the operated and the adjacent segments before and after X-Stop surgeries during various weight-bearing postures in elderly patients with lumbar spine stenosis. The mechanism of interspinous process (ISP) devices is to directly distract the ISP of the implanted level to indirectly decompress the intervertebra foramen and spinal canal. Few studies have investigated the changes of ISP gap caused by X-Stop implantation using magnetic resonance imaging or radiography, but the effect of X-Stop surgery on the kinematics of spinous processes during functional activities is still unclear. Eight patients were tested before and, on average, 7 months after surgical implantation of the X-Stop devices using a combined computed tomography/magnetic resonance imaging and dual fluoroscopic imaging system during weight-bearing standing, flexion-extension, left-right bending, and left-right twisting positions of the torso. The shortest distances of the ISPs at the operated and the adjacent levels were measured using iterative closest point method and was dissected into vertical (gap) and horizontal (lateral translation) components. At the operated levels, the shortest vertical ISP distances (gap) significantly (P<0.05) increased by 1.5 mm during standing, 1.2 mm during left twist, 1.3 mm during extension, and 1.1 mm during flexion, whereas they also increased yet not significantly (P>0.05) in right twist, left bend, and right bend after the X-Stop implantation. The lateral translations were not significantly affected. At both cephalad and caudad adjacent levels, the ISP distances (vertical and horizontal) were not significantly affected during all postures after X-Stop implantation. The findings of this study indicate that implantation of the X-Stop devices can effectively distract the ISP space at the diseased level without causing apparent kinematic changes at the adjacent segments during the studied postures.

Weighted ICP Algorithm for Alignment of Stars from Scanned Astronomical Photographic Plates

Given the coarse celestial coordinates of the centre of a plate scan and the field of view, we are looking for a mapping between the stars extracted from the image and the stars from a catalogue, where the stars from both sources are represented by their stellar magnitudes and coordinates, relatively to the image centre. In a previous work we demonstrated the application of Iterative Closest Point (ICP) algorithm for the alignment problem where stars were represented only by their geometrical coordinates. ICP leads to translation and rotation of the initial points - a correction required for one set of stars to fit over the other. This paper extends the previous work by demonstrating significant improvement of ICP by using the stellar magnitudes as point weights. The improvement consists of great decrease of the iteration count until convergence, which helps in the case of highly “misaligned” initial states. The essential aspects of the ICP method like noise tolerance of false or missing stars are still in charge.

On the detection of systematic errors in terrestrial laser scanning data

Quality descriptions are parts of the key tasks of geodetic data processing. Systematic errors should be detected and avoided in order to insure the high quality standards required by structural monitoring. In this study, the iterative closest point (ICP) method was invested to detect systematic errors in two overlapping data sets. There are three steps to process the systematic errors: firstly, one of the data sets was transformed to a reference system by the introduction of the Gauss–Helmert (GH) model. Secondly, quadratic form estimation and segmentation methods are proposed to guarantee the overlapping data sets. Thirdly, the ICP method was employed for a finer registration and detecting the systematic errors. A case study was casted in which a dam surface in Germany was scanned by terrestrial laser scanning (TLS) technology. The results indicated that with the conjugation of ICP algorithm the accuracy of the data sets was improved approximately by 1.6 mm.

Adjustment of systematic errors in ALS data through surface matching

Surface matching is a well researched topic in both Computer Vision (CV) and terrestrial laser scanning (TLS) or ground based light detection and ranging (LiDAR), but the extent of the range images derived from these technologies is typically orders of magnitude smaller than those derived from airborne laser scanning (ALS), also known as airborne LiDAR. Iterative closest point (ICP) and its variants have been successfully used to align and register multiple overlapping views of the range images for CV and TLS applications. However, many challenges are encountered in applying the ICP approach to ALS data sets. In this paper, we address these issues, explore the possibility of automating the algorithm, and present a technique to adjust systematic discrepancies in overlapping strips, using geometrical attributes in a given terrain. In this method, the ALS point samples used in the algorithm are selected depending on their ability to constrain the relative movement between the overlapping laser strips. The points from overlapping strips are matched through modified point to plane based on the ICP method.

Bone Alignment Using the Iterative Closest Point Algorithm

Computer assisted surgical interventions and research in joint kinematics rely heavily on the accurate registration of three-dimensional bone surface models reconstructed from various imaging technologies. Anomalous results were seen in a kinematic study of carpal bones using a principal axes alignment approach for the registration. The study was repeated using an iterative closest point algorithm, which is more accurate, but also more demanding to apply. The principal axes method showed errors between 0.35 mm and 0.49 mm for the scaphoid, and between 0.40 mm and 1.22 mm for the pisiform. The iterative closest point method produced errors of less than 0.4 mm. These results show that while the principal axes method approached the accuracy of the iterative closest point algorithm in asymmetrical bones, there were more pronounced errors in bones with some symmetry. Principal axes registration for carpal bones should be avoided.

A Novel Kernel Correlation Model with the Correspondence Estimation

We present a novel multiple-linked iterative closest point method to estimate correspondences and the rigid/non-rigid transformations between point-sets or shapes. The estimation task is carried out by maximizing a symmetric similarity function, which is the product of the square roots of correspondences and a kernel correlation. The local mean square error analysis and robustness analysis are provided to show our method's superior performance to the kernel correlation method.