3D Matching Research Articles

SU‐E‐I‐58: Detecting Tumors with Extremely Low Contrast in CT Images

Purpose:Tumors such as the prostate focal lesions and the brain metastases have extremely low CT contrast and MRI is usually used for target delineation. The target contours are propagated to the CT for treatment planning and patient positioning. We have employed an advanced denoising method eliminating the noise and allow magnification of subtle contrast of these focal lesions on CT.Methods:Five prostate and two brain metastasis patients with MRI T2, diffusion or dynamic contrast enhanced (DCE) images confirmed focal lesions were included. One brain patients had 5 metastases. A block matching 3D (BM3D) algorithm was adapted to reduce the noise of kVCT images used for treatment planning. The gray‐level range of the resultant images was narrowed to magnify the tumor‐normal tissue contrast.Results:For the prostate patients, denoised kVCT images showed focal regions at 5, 8,11‐1, 2, and 8–10 oclock for the 5 patients, this is highly consistent to the radiologist confirmed focal lesions based on MRI at 5, 7, 11‐1, 2 and 8–10 oclock in the axial plane. These CT focal regions matched well with the MRI focal lesions in the cranio‐caudal position. The average increase in density compared to background prostate glands was 0.86%, which corresponds to ∼50% increase in cellularity and is lower than the average CT noise level of 2.4%. For the brain patients, denoised kVCT showed 5/6 metastases. The high CT‐density region of a metastasis is 2‐mm off from its corresponding elevated MRI perfusion center. Overall the detecting sensitivity was 91%.Conclusion:It has been preliminarily demonstrated that the higher tumor cellularity can be detected using kVCT. The low contrast‐to‐noise information requires advanced denoising to reveal. The finding is significant to radiotherapy by providing an additional tool to locate focal lesions for confirming MRI‐CT registration and providing a highly accessible outcome assessment tool.

Target visibility enhancement for C-arm cone beam CT-fluoroscopy-guided hepatic needle placement: implementation and accuracy evaluation

Fluoroscopy-guided hepatic intervention is limited by target visibility and respiratory movement. A feasible procedure for visibility enhancement of the key regions and targets in 2D fluoroscopic images is needed. A system was developed to improve targeting by integrating the forward projection of objects extracted from 3D cone beam CT (CBCT) volumes. The target matching accuracy during regular respiration was measured to evaluate the system. 3D CBCT abdominal volumes were acquired and segmented to extract different regions, including the diaphragm, hepatic vessels, bony structures, and hepatic tumor. The segmented result was rendered and projected to generate augmented fluoroscopy fusion images. The target matching accuracy by applying these procedures was evaluated for the hepatic intervention guidance. Quantitative assessment of the target matching accuracy in the upper section of liver was performed for eight targets from four subjects. The 2D and 3D target matching accuracy were 0.98±0.37 and 1.47±0.26 mm, respectively. The 2D target matching accuracy was 1.46±0.67 mm for the target in the lower liver. This accuracy should be acceptable for the 5 mm safety margin required in clinical use. Visibility of targets in 2D fluoroscopy was enhanced to improve interactive navigation guidance for hepatic needle placement. The target matching accuracy for the C-arm cone beam CT-fluoroscopy-guided hepatic needle targeting was sufficient for clinical use.

Image-based correlation of Laser Scanning point cloud time series for landslide monitoring

Very high resolution monitoring of landslide kinematics is an important aspect for a physical understanding of the failure mechanisms and for quantifying the associated hazard. In the last decade, the potential of Terrestrial Laser Scanning (TLS) to monitor slow-moving landslides has been largely demonstrated but accurate processing methods are still needed to extract useful information available in point cloud time series. This work presents an approach to measure the 3D deformation and displacement patterns from repeated TLS surveys. The method is based on the simplification of a 3D matching problem in a 2D matching problem by using a 2D statistical normalized cross-correlation function. The computed displacement amplitudes are compared to displacements (1) calculated with the classical approach of Iterative Closest Point and (2) measured from repeated dGPS observations. The performance of the method is tested on a 3 years dataset acquired at the Super-Sauze landslide (South French Alps). The observed landslide displacements are heterogeneous in time and space. Within the landslide, sub-areas presenting different deformation patterns (extension, compression) are detected by a strain analysis. It is demonstrated that pore water pressure changes within the landslide is the main controlling factor of the kinematics.

3D ear identification based on sparse representation.

Biometrics based personal authentication is an effective way for automatically recognizing, with a high confidence, a person’s identity. Recently, 3D ear shape has attracted tremendous interests in research field due to its richness of feature and ease of acquisition. However, the existing ICP (Iterative Closet Point)-based 3D ear matching methods prevalent in the literature are not quite efficient to cope with the one-to-many identification case. In this paper, we aim to fill this gap by proposing a novel effective fully automatic 3D ear identification system. We at first propose an accurate and efficient template-based ear detection method. By utilizing such a method, the extracted ear regions are represented in a common canonical coordinate system determined by the ear contour template, which facilitates much the following stages of feature extraction and classification. For each extracted 3D ear, a feature vector is generated as its representation by making use of a PCA-based local feature descriptor. At the stage of classification, we resort to the sparse representation based classification approach, which actually solves an l1-minimization problem. To the best of our knowledge, this is the first work introducing the sparse representation framework into the field of 3D ear identification. Extensive experiments conducted on a benchmark dataset corroborate the effectiveness and efficiency of the proposed approach. The associated Matlab source code and the evaluation results have been made publicly online available at http://sse.tongji.edu.cn/linzhang/ear/srcear/srcear.htm.

3D soft-tissue tracking using spatial-color joint probability distribution and thin-plate spline model

Visual tracking techniques based on stereo endoscope are developed to measure tissue motion in robot-assisted minimally invasive surgery. However, accurate 3D tracking of tissue surfaces remains challenging due to complicated deformation, poor imaging conditions, specular reflections and other dynamic effects during surgery. This study employs a robust and efficient 3D tracking scheme with two independent recursive processes, namely kernel-based inter-frame motion estimation and model-based intra-frame 3D matching. In the first process, target region is represented in joint spatial-color space for robust estimation. By defining a probabilistic similarity measure, a mean-shift-based iterative algorithm is derived for location of the target region in a new image. In the second process, the thin-plate spline model is used to fit the 3D shape of tissue surfaces around the target region. An iterative algorithm based on an efficient second-order minimization technique is derived to compute optimal model parameters. The two processes can be computed in parallel. Their outputs are combined to recover 3D information about the target region. The performance of the proposed method is validated using phantom heart videos and in vivo videos acquired by the daVinci® surgical robotic platform and a synthesized data set with known ground truth.

3D ear recognition using local salience and principal manifold

As an emerging class of biometrics, human ear has drawn significant attention in recent years. In this paper, we propose a novel 3D ear shape matching and recognition system. First, we propose a novel method for computing saliency value of each point on 3D ear point clouds, which is based on the Gaussian-weighted average of the mean curvature and can be used to sort the keypoints accordingly. Then we propose the optimal selection of the salient key points using the Poisson Disk Sampling. Finally, we fit a surface to the neighborhood of each salient keypoint using the quadratic principal manifold method, establishing the local feature descriptor of each salient keypoint. The experimental results on ear shape matching show that, compared with other similar methods, the proposed system has higher approximation precision on shape feature detection and higher matching accuracy on the ear recognition.

Fast glacier volume change detection based on least squares 3D surface matching of DEM

In this paper, a new fast detection method on glacier volume changing has been adopted and analyzed to solve shortages of low precision and inefficiency in traditional algorithms. High-precision Digital Elevation Model (DEM) has been used in this method to improve the accuracy of the result. Besides, GCPs are unnecessary in this new method so that it can be widely applied in glacier districts where GCPs are hard to be set. Furthermore, Least squares 3D surface matching method has also been improved in this paper in that blocking computing has been added to strengthen the speed and accuracy of this method. Multitemporal images in Xinjiang Province, china have been used to test the accuracy and efficiency of this proposal method: iterations are 12 and D-values are small. The test results show that this method has the advantages of Least squares 3D surface matching so that it can be widely used to satisfy the speed and accuracy of fast changing detection of Glacier Volume.

Digital Aléria: Multiresolution Technologies Arming International Collaboration to Restore the Past and Restart

This paper reports about the last progressing results of our work, raised around the archaeological site of the ancient roman urbs of Aléria in Corsica (F), updated to January 2014. Our plan provides to resume the old excavation usable data, to realize a new survey of the site actual status and to collect all into a new 3D geodatabase, including any kind of historical, documental analog data, all reorganized into dedicated Cloud Platforms and Services for different uses, as a multipath solution to improve scientists' work and to promote tourism, both local and international. The project initially focused on the forum, as the only accessible area, in these days enlarges its plans involving the nearby amphitheater and ramparts, just reorganized after the removal of old roofing. It somehow embraces even the nearby Carcopino Museum and the collected Etruscan and Roman finds as traces of the unruly diggings previously realized, to which we'll attempt to restitute logic and documented belonging. Since the Aléria Carcopino Museum shows important archaeological evidences sharing the same origin with some finds exposed in the National Etruscan Museum of Villa Giulia in Rome, a new initiative was set up in these days by surveying the two museums' buildings, to create 3D georeferenced models as containers for sharing digital multiscale data made by innovative computational photography techniques there applied. This operation is now in place in the Villa Giulia Museum as a prototype to be after web-linked to the same survey model to be realized in Aléria. Leading the whole project is the collaboration agreement between the “Collectivité Territoriale de Corse, secteur Archéologie”, and the Department of “Storia, Disegno e Restauro dell'Architettura” SAPIENZA, Italy, formally signed in year 2013. The process employs many multiresolution surveying technologies: 3D Laser scanning, Topography, GPS and Dense Stereo Matching applied on site, on buildings and objects for manifold purposes; moreover Computational Photography techniques applied on Museum's halls, cabinets and exposed objects, to create digital and interactive exhibitions linking both nations. A GIS and WEBGIS workflow followed, to start collecting new and restored databases to build up a complete 3D geo-database with 3D Web scene to be shared in the GIS online Cloud Platform.

Midsagittal plane extraction from brain images based on 3D SIFT

Midsagittal plane (MSP) extraction from 3D brain images is considered as a promising technique for human brain symmetry analysis. In this paper, we present a fast and robust MSP extraction method based on 3D scale-invariant feature transform (SIFT). Unlike the existing brain MSP extraction methods, which mainly rely on the gray similarity, 3D edge registration or parameterized surface matching to determine the fissure plane, our proposed method is based on distinctive 3D SIFT features, in which the fissure plane is determined by parallel 3D SIFT matching and iterative least-median of squares plane regression. By considering the relative scales, orientations and flipped descriptors between two 3D SIFT features, we propose a novel metric to measure the symmetry magnitude for 3D SIFT features. By clustering and indexing the extracted SIFT features using a k-dimensional tree (KD-tree) implemented on graphics processing units, we can match multiple pairs of 3D SIFT features in parallel and solve the optimal MSP on-the-fly. The proposed method is evaluated by synthetic and in vivo datasets, of normal and pathological cases, and validated by comparisons with the state-of-the-art methods. Experimental results demonstrated that our method has achieved a real-time performance with better accuracy yielding an average yaw angle error below 0.91° and an average roll angle error no more than 0.89°.

Recognizing activities in multiple views with fusion of frame judgments

This paper focuses on activity recognition when multiple views are available. In the literature, this is often performed using two different approaches. In the first one, the systems build a 3D reconstruction and match that. However, there are practical disadvantages to this methodology since a sufficient number of overlapping views is needed to reconstruct, and one must calibrate the cameras. A simpler alternative is to match the frames individually. This offers significant advantages in the system architecture (e.g., it is easy to incorporate new features and camera dropouts can be tolerated). In this paper, the second approach is employed and a novel fusion method is proposed. Our fusion method collects the activity labels over frames and cameras, and then fuses activity judgments as the sequence label. It is shown that there is no performance penalty when a straightforward weighted voting scheme is used. In particular, when there are enough overlapping views to generate a volumetric reconstruction, our recognition performance is comparable with that produced by volumetric reconstructions. However, if the overlapping views are not adequate, the performance degrades fairly gracefully, even in cases where test and training views do not overlap.

Fast blockwise SURE shrinkage for image denoising

In this paper, we investigate the shrinkage problem of image denoising for various methods under the additive white Gaussian noise (AWGN) model. Our main contribution is to derive the closed-form of the optimal shrinkage that minimizes the Stein's unbiased risk estimator (SURE) and thus allows direct blockwise shrinkage without additional optimizations. Simulation results show that the proposed method boosts the denoising performance for a variety of image denoising techniques including the moving average filter, the median filter, the wiener filter, the bilateral filter, the probabilistic non-local means, and the block matching 3D filter in terms of higher pixel signal noise ratio (PSNR) and structural similarity index (SSIM). We also case study the proposed shrinkage solution with respect to the classic NLM denoising, whose shrinkage solutions and equivalent forms have been widely researched, and further confirm its superiority. The proposed shrinkage solution can be used to improve arbitrary image denoising methods under the AWGN model, and it serves as a good remedy to save badly denoised images due to inappropriate parameters.

Detection of the Single Image from DIBR Based on 3D Warping Trace and Edge Matching

Recently, the popularity of 3D content is on the rise because of its immersive experience to view- ers. While demands for 3D contents and 3D technologies increase, only a few copyright protection methods for 3D contents have been proposed. The simplest infringement is the illegal distribution of the single 2D image from 3D content. The leaked image is still valuable as 2D content and the leakage can be occurred in DIBR system. To detect the leaked image, we focus on the hole-filled region which is caused by the hole-filling procedure mandatory in DIBR system. To estimate the hole-filled regions, two different procedures are conducted to extract edges and to estimate 3D warping traces, respectively. After that, the hole-filled regions are estimated and the left-right-eye image discrimination (LR discrimination) is also conducted. Experimental results demonstrate the effectiveness of the proposed method using quantitative measures.

3D Face Matching Based on Depth-Level Curves

In the domain of 3D face matching, many techniques have been developed as variants of 3D facial matching approaches that reduce the amount of facial data into few 3D curves. In the literature, many curves have been considered: level-curves, radial curves, iso-stripes, crest lines, etc... In this paper, we exploit curve concept to represent 3D facial geometric. First, depth-level curves were extracted to present 3D facial data then, we investigate the dimensionality reduction offered by Random Projection to perform an artificial system for face recognition using Back-propagation neural network. Experiment was conducted using vrml files from FRAV Database considering only one training sample per person.

3D foot and shoe matching based on OBB and AABB

Purpose – The purpose of this paper is to present a novel method of 3D foot and shoe model matching based on oriented bounding box (OBB) and axis-aligned bounding box (AABB). Design/methodology/approach – The paper first calculates their OBBs of foot and shoe models; aligns three axial directions of their OBBs to be parallel to three axes of world coordinate system. Then, computes their AABBs of foot and shoe models, translates the center of the bottom face of the foot's AABB to that of the shoe's AABB. Findings – After the matching, the shoe model could be larger or locally smaller than the foot model. The paper finally adjusts the size of shoe model according to the distance difference. Originality/value – Experimental results show that this method is simple and feasible which can effectively realize the matching between foot and shoe models.

Adaptive multiple subtraction based on 3D blind separation of convolved mixtures

Multiple removal is one of the important preprocessing steps in a seismic data processing sequence. For an accurate multiple prediction, a full 3D method is required. However, the application of such methods is often limited by practical and economic constraints. Therefore, in practical situations, 2D prediction methods are used, such as 2D surface-related multiple elimination. However, the resulting predicted multiples may have significant temporal shift, spatial mismatch, and amplitude inconsistency, compared with true 3D multiples. Adaptive multiple subtraction based on 2D blind separation of convolved mixtures (BSCM) has been proposed to estimate the 2D matching filter in a single gather. To improve the flexibility of the adaptive multiple subtraction for the inconsistencies between the 2D predicted multiples and true 3D multiples, we evaluated the adaptive multiple subtraction as a problem of 3D BSCM. In the proposed method, the predicted multiples were modeled as the convolution of the true multiples with a 3D kernel, whose third dimension is in the gather direction. By maximizing the non-Gaussianity of the estimated primaries, the iterative reweighted least-squares algorithm was exploited to obtain the 3D matching filter, which is the inverse of the 3D kernel. To avoid the possible overfitting problem introduced by the 3D matching filter, the proposed method fit several seismic gathers using one 3D matching filter. In addition, by using the non-Gaussian maximization criterion, the proposed method alleviated the orthogonality assumption used by the least-squares subtraction method. Furthermore, the proposed method can eliminate the temporal and spatial mismatches between the 2D predicted multiples and true 3D multiples better than the 2D BSCM subtraction method. Tests on synthetic and field data sets demonstrated the effectiveness of the 3D BSCM subtraction method.

Detailed 3D Representations for Object Recognition and Modeling

Geometric 3D reasoning at the level of objects has received renewed attention recently in the context of visual scene understanding. The level of geometric detail, however, is typically limited to qualitative representations or coarse boxes. This is linked to the fact that today's object class detectors are tuned toward robust 2D matching rather than accurate 3D geometry, encouraged by bounding-box-based benchmarks such as Pascal VOC. In this paper, we revisit ideas from the early days of computer vision, namely, detailed, 3D geometric object class representations for recognition. These representations can recover geometrically far more accurate object hypotheses than just bounding boxes, including continuous estimates of object pose and 3D wireframes with relative 3D positions of object parts. In combination with robust techniques for shape description and inference, we outperform state-of-the-art results in monocular 3D pose estimation. In a series of experiments, we analyze our approach in detail and demonstrate novel applications enabled by such an object class representation, such as fine-grained categorization of cars and bicycles, according to their 3D geometry, and ultrawide baseline matching.

DELINEATION OF BUILDING FOOTPRINTS FROM HIGH RESOLUTION SATELLITE STEREO IMAGERY USING IMAGE MATCHING AND A GIS DATABASE

Abstract. In this paper, a workflow is proposed to delineate building footprints from high resolution satellite stereo images through integration of a Digital Surface Model (DSM), 3D edge matching technique and GIS building polygons. First, Digital Surface Models (DSM) and a normalised DSM are derived by traditional image matching. Three different removal masks are employed to reduce the effects of matching errors, roads and vegetation. We then compare all resulting blobs with the GIS 2D building layer. For blobs without a partner in the database, building outlines are extracted based on 3D edge matching. To do so, edges on epipolar images are first detected individually on each image using the Canny operator. After removing short edges and extracting straight lines, we find the best corresponding lines using various constraints. Finally, the topological relationship of the derived 3D edges is employed to reconstruct the shape of the building and the building footprints using a box-fitting approach. Our experiments on GeoEye-1 stereo images show promising results if buildings are large enough, have simple shape and show good contrast compared to the background.

Optimizing SIFT for Matching of Short Wave Infrared and Visible Wavelength Images

The scale invariant feature transform (SIFT) is a widely used interest operator for supporting tasks such as 3D matching, 3D scene reconstruction, panorama stitching, image registration and motion tracking. Although SIFT is reported to be robust to disparate radiometric and geometric conditions in visible light imagery, using the default input parameters does not yield satisfactory results when matching imagery acquired at non-overlapping wavelengths. In this paper, optimization of the SIFT parameters for matching multi-wavelength image sets is documented. In order to integrate hyperspectral panoramic images with reference imagery and 3D data, corresponding points were required between visible light and short wave infrared images, each acquired from a slightly different position and with different resolutions and geometric projections. The default SIFT parameters resulted in too few points being found, requiring the influence of five key parameters on the number of matched points to be explored using statistical techniques. Results are discussed for two geological datasets. Using the SIFT operator with optimized parameters and an additional outlier elimination method, allowed between four and 22 times more homologous points to be found with improved image point distributions, than using the default parameter values recommended in the literature.

Speckle-constrained variational methods for image restoration in optical coherence tomography

A number of despeckling methods for optical coherence tomography (OCT) have been proposed. In these digital filtering techniques, speckle noise is often simplified as additive white Gaussian noise due to the logarithmic compression for the signal. The approximation is not completely consistent with the characteristic of OCT speckle noise, and cannot be reasonably extended to deconvolution algorithms. This paper presents a deconvolution model that combines the variational regularization term with the statistical characteristic constraints of data corrupted by OCT speckle noise. In the data fidelity term, speckle noise is modeled as signal dependent, and the point spread function of OCT systems is included. The regularization functional introduces a priori information on the original images, and a regularization term based on block matching 3D modeling is used to construct the variational model in the paper. Finally, the method is applied to the restoration of actual OCT raw data of human skin. The numerical results demonstrate that the proposed deconvolution algorithm can simultaneously enhance regions of images containing detail and remove OCT speckle noise.

Multiscale 3D feature extraction and matching with an application to 3D face recognition

We present a new multiscale surface representation for 3D shape matching that is based on scale-space theory. The representation, Curvature Scale-Space 3D (CS3), is well-suited for measuring dissimilarity between (partial) surfaces having unknown position, orientation, and scale. The CS3 representation is obtained by evolving the surface curvatures according to the heat equation. This evolution process yields a stack of increasingly smoothed surface curvatures that is useful for keypoint extraction and descriptor computations. We augment this information with an associated scale parameter at each stack level to define our multiscale CS3 surface representation. The scale parameter is necessary for automatic scale selection, which has proven to be successful in 2D scale-invariant shape matching applications. We show that our keypoint and descriptor computation approach outperforms many of the leading methods. The main advantages of our representation are its computational efficiency, lower memory requirements, and ease of implementation.