Showing papers on "RANSAC published in 2008"

Active Matching

Abstract: In the matching tasks which form an integral part of all types of tracking and geometrical vision, there are invariably priors available on the absolute and/or relative image locations of features of interest. Usually, these priors are used post-hoc in the process of resolving feature matches and obtaining final scene estimates, via `first get candidate matches, then resolve' consensus algorithms such as RANSAC. In this paper we show that the dramatically different approach of using priors dynamically to guide a feature by feature matching search can achieve global matching with much fewer image processing operations and lower overall computational cost. Essentially, we put image processing into the loopof the search for global consensus. In particular, our approach is able to cope with significant image ambiguity thanks to a dynamic mixture of Gaussians treatment. In our fully Bayesian algorithm, the choice of the most efficient search action at each step is guided intuitively and rigorously by expected Shannon information gain. We demonstrate the algorithm in feature matching as part of a sequential SLAM system for 3D camera tracking. Robust, real-time matching can be achieved even in the previously unmanageable case of jerky, rapid motion necessitating weak motion modelling and large search regions.

Real time detection of lane markers in urban streets

Abstract: We present a robust and real time approach to lane marker detection in urban streets. It is based on generating a top view of the road, filtering using selective oriented Gaussian filters, using RANSAC line fitting to give initial guesses to a new and fast RANSAC algorithm for fitting Bezier Splines, which is then followed by a post-processing step. Our algorithm can detect all lanes in still images of the street in various conditions, while operating at a rate of 50 Hz and achieving comparable results to previous techniques.

A Comparative Analysis of RANSAC Techniques Leading to Adaptive Real-Time Random Sample Consensus

Abstract: The Random Sample Consensus (RANSAC) algorithm is a popular tool for robust estimation problems in computer vision, primarily due to its ability to tolerate a tremendous fraction of outliers. There have been a number of recent efforts that aim to increase the efficiency of the standard RANSAC algorithm. Relatively fewer efforts, however, have been directed towards formulating RANSAC in a manner that is suitable for real-time implementation. The contributions of this work are two-fold: First, we provide a comparative analysis of the state-of-the-art RANSAC algorithms and categorize the various approaches. Second, we develop a powerful new framework for real-time robust estimation. The technique we develop is capable of efficiently adapting to the constraints presented by a fixed time budget, while at the same time providing accurate estimation over a wide range of inlier ratios. The method shows significant improvements in accuracy and speed over existing techniques.

Optimal Randomized RANSAC

Abstract: A randomized model verification strategy for RANSAC is presented. The proposed method finds, like RANSAC, a solution that is optimal with user-specified probability. The solution is found in time that is close to the shortest possible and superior to any deterministic verification strategy. A provably fastest model verification strategy is designed for the (theoretical) situation when the contamination of data by outliers is known. In this case, the algorithm is the fastest possible (on the average) of all randomized RANSAC algorithms guaranteeing a confidence in the solution. The derivation of the optimality property is based on Wald's theory of sequential decision making, in particular, a modified sequential probability ratio test (SPRT). Next, the R-RANSAC with SPRT algorithm is introduced. The algorithm removes the requirement for a priori knowledge of the fraction of outliers and estimates the quantity online. We show experimentally that on standard test data, the method has performance close to the theoretically optimal and is 2 to 10 times faster than standard RANSAC and is up to four times faster than previously published methods.

Multiframe Motion Segmentation with Missing Data Using PowerFactorization and GPCA

Abstract: We consider the problem of segmenting multiple rigid-body motions from point correspondences in multiple affine views. We cast this problem as a subspace clustering problem in which point trajectories associated with each motion live in a linear subspace of dimension two, three or four. Our algorithm involves projecting all point trajectories onto a 5-dimensional subspace using the SVD, the PowerFactorization method, or RANSAC, and fitting multiple linear subspaces representing different rigid-body motions to the points in ?5 using GPCA. Unlike previous work, our approach does not restrict the motion subspaces to be four-dimensional and independent. Instead, it deals gracefully with all the spectrum of possible affine motions: from two-dimensional and partially dependent to four-dimensional and fully independent. Our algorithm can handle the case of missing data, meaning that point tracks do not have to be visible in all images, by using the PowerFactorization method to project the data. In addition, our method can handle outlying trajectories by using RANSAC to perform the projection. We compare our approach to other methods on a database of 167 motion sequences with full motions, independent motions, degenerate motions, partially dependent motions, missing data, outliers, etc. On motion sequences with complete data our method achieves a misclassification error of less that 5% for two motions and 29% for three motions.

A method for automated registration of unorganised point clouds

Abstract: A registration method for unorganised point cloud datasets, the Geometric Primitive ICP with the RANSAC (GP-ICPR), which uses geometric primitives, neighbourhood search and the positional uncertainty of laser scanners is proposed. The change of geometric curvature and approximate normal vector of the surface formed by a point and its neighbourhood are used to search for possible corresponding points. The GP-ICPR was tested with terrestrial laser scanner datasets in terms of its precision and accuracy. It is shown that the GP-ICPR improved the precision of the estimated relative transformation parameters by as much as a factor of 5, which provides a window of opportunity to utilise this automated registration method in practical applications such as terrestrial surveying and deformation monitoring.

Extended ransac algorithm for automatic detection of building roof planes from lidar data

Abstract: Airborne laser scanner technique is broadly the most appropriate way to acquire rapidly and with high density 3D data over a city. Once the 3D lidar data are available, the next task is the automatic data processing, with major aim to construct 3D building models. Among the numerous automatic reconstruction methods, the techniques allowing the detection of 3D building roof planes are of crucial importance. For this purpose, this paper studies the Random Sample Consensus (RANSAC) algorithm. Its principle and pseudocode - seldom detailed in the related literature - as well as its complete analyse are presented in this paper. Despite all advantages of this algorithm, it gives sometimes erroneous results. That can be explained by the fact that it uses a pure mathematical principle for detecting the roof planes. So it looks for the best plane without taking into account the particularity of the captured object. The extended RANSAC algorithm proposed in this paper allows harmonizing the mathematical aspect of the algorithm with the geometry of a roof. It is shown that the extended approach provides very satisfying results, even in the case of very weak point density and for different levels of building complexity. Moreover, the adjacency relationships of the neighbouring roof planes are described and analysed. Hence the roof planes are successfully detected and adjacency relationships of the adjacent roof planes are calculate. Finally the automatic building modelling can be carried out easily.

Robust Gradient-Based 3-D/2-D Registration of CT and MR to X-Ray Images

Abstract: One of the most important technical challenges in image-guided intervention is to obtain a precise transformation between the intrainterventional patient's anatomy and corresponding preinterventional 3-D image on which the intervention was planned. This goal can be achieved by acquiring intrainterventional 2-D images and matching them to the preinterventional 3-D image via 3-D/2-D image registration. A novel 3-D/2-D registration method is proposed in this paper. The method is based on robustly matching 3-D preinterventional image gradients and coarsely reconstructed 3-D gradients from the intrainterventional 2-D images. To improve the robustness of finding the correspondences between the two sets of gradients, hypothetical correspondences are searched for along normals to anatomical structures in 3-D images, while the final correspondences are established in an iterative process, combining the robust random sample consensus algorithm (RANSAC) and a special gradient matching criterion function. The proposed method was evaluated using the publicly available standardized evaluation methodology for 3-D/2-D registration, consisting of 3-D rotational X-ray, computed tomography, magnetic resonance (MR), and 2-D X-ray images of two spine segments, and standardized evaluation criteria. In this way, the proposed method could be objectively compared to the intensity, gradient, and reconstruction-based registration methods. The obtained results indicate that the proposed method performs favorably both in terms of registration accuracy and robustness. The method is especially superior when just a few X-ray images and when MR preinterventional images are used for registration, which are important advantages for many clinical applications.

A polynomial-time bound for matching and registration with outliers

Abstract: We present a framework for computing optimal transformations, aligning one point set to another, in the presence of outliers. Example applications include shape matching and registration (using, for example, similarity, affine or projective transformations) as well as multiview reconstruction problems (triangulation, camera pose etc.). While standard methods like RANSAC essentially use heuristics to cope with outliers, we seek to find the largest possible subset of consistent correspondences and the globally optimal transformation aligning the point sets. Based on theory from computational geometry, we show that this is indeed possible to accomplish in polynomial-time. We develop several algorithms which make efficient use of convex programming. The scheme has been tested and evaluated on both synthetic and real data for several applications.

An Efficient Approach to Onboard Stereo Vision System Pose Estimation

Abstract: This paper presents an efficient technique for estimating the pose of an onboard stereo vision system relative to the environment's dominant surface area, which is supposed to be the road surface. Unlike previous approaches, it can be used either for urban or highway scenarios since it is not based on a specific visual traffic feature extraction but on 3D raw data points. The whole process is performed in the Euclidean space and consists of two stages. Initially, a compact 2D representation of the original 3D data points is computed. Then, a RANdom SAmple Consensus (RANSAC) based least-squares approach is used to fit a plane to the road. Fast RANSAC fitting is obtained by selecting points according to a probability function that takes into account the density of points at a given depth. Finally, stereo camera height and pitch angle are computed related to the fitted road plane. The proposed technique is intended to be used in driver-assistance systems for applications such as vehicle or pedestrian detection. Experimental results on urban environments, which are the most challenging scenarios (i.e., flat/uphill/downhill driving, speed bumps, and car's accelerations), are presented. These results are validated with manually annotated ground truth. Additionally, comparisons with previous works are presented to show the improvements in the central processing unit processing time, as well as in the accuracy of the obtained results.

Integrated speed limit detection and recognition from real-time video

Abstract: Here we propose a complete system for robust detection and recognition of the current speed sign restrictions from a moving road vehicle. This approach includes the detection and recognition of both numerical limit and national limit (cancellation) signs with the addition of automatic vehicle turn detection. The system utilizes both RANSAC-based colour-shape detection of speed limit signs and neural network based recognition whilst turn analysis relies on an optic flow based method. As primary detection is based on a robust colour and shape detection methodology this results in a real-time algorithm that is invariant to variable road conditions. The integration of both limit, cancellation and vehicle turn detection within the bounds of real-time system performance represents an advance on prior work within this field.

Method and system for real-time visual odometry

Abstract: A method for real-time visual odometry comprises capturing a first three-dimensional image of a location at a first time, capturing a second three-dimensional image of the location at a second time that is later than the first time, and extracting one or more features and their descriptors from each of the first and second three-dimensional images. One or more features from the first three-dimensional image are then matched with one or more features from the second three-dimensional image. The method further comprises determining changes in rotation and translation between the first and second three-dimensional images from the first time to the second time using a random sample consensus (RANSAC) process and a unique iterative refinement technique.

Curb Detection Based on a Multi-Frame Persistence Map for Urban Driving Scenarios

Abstract: An approach for the detection of straight and curved curbs (border of relevant traffic isles, sidewalks, etc) is presented, in the context of urban driving assistance systems. A rectangular elevation map is built from 3D dense stereo data. Edge detection is applied to the elevation map in order to highlight height variations. We propose a method to reduce significantly the 3D noise from dense stereo, using a multiframe persistence me persistence map: temporal filtering is performed for edge points, based on the ego car motion, and only persistent points are validated. The Hough accumulator for lines is built with the persistent edge points. A scheme for extracting straight curbs (as curb segments) and curved curbs (as chains of curb segments) is proposed. Each curb segment is refined using a RANSAC approach to fit optimally the 3D data of the curb. The algorithm was evaluated in an urban scenario. It works in real-time and provides robust detection of curbs.

Automatic image stitching using SIFT

Abstract: This paper concerns the problem of automatic image stitching which mainly applies to the image sequence even those including noise images. And it uses a method based on invariant features to realize fully automatic image stitching, in which it includes two main parts: image matching and image blending. As the noises images have large differences between the other images, when using SIFT features to realize correct and robust matching, it supplies a probabilistic model to verify the panorama image sequence. Addison to have a more satisfied panorama image, it uses a simple and fast blending method which is weighted average method. Finally, the experiment results confirm the feasibility of our methods.

A Stairway Detection Algorithm based on Vision for UGV Stair Climbing

Abstract: In the paper, we present a vision based algorithm used to guide the unmanned ground vehicles (UGV) for autonomous stairways climbing and implement it on UGV successfully. The reliability of guiding UGV to climb stairs requires evaluating two offset parameters: the position of vehicle on stairs and the orientation angle to stairs. The intention of our algorithm is to estimate these two parameters through extracting the stair edges robustly. To achieve this goal, we apply the Gabor filter to eliminate the influence of the illumination and keep edges, and propose a fast method to remove small lines. Finally we link stair edges, and estimate the offset parameters used to steer the vehicle by RANSAC algorithm. Experiments on various stairways including indoor and outdoor are given in various light conditions. The results validate our algorithm.

Robust Optimal Pose Estimation

Abstract: We study the problem of estimating the position and orientationof a calibrated camera from an image of a known scene. A commonproblem in camera pose estimation is the existence of falsecorrespondences between image features and modeled 3D points.Existing techniques such as RANSAC to handle outliers have noguarantee of optimality. In contrast, we work with a naturalextension of the L∞ norm to the outlier case. Usinga simple result from classical geometry, we derive necessaryconditions for L∞ optimality and show how to usethem in a branch and bound setting to find the optimum and todetect outliers. The algorithm has been evaluated on synthetic aswell as real data showing good empirical performance. In addition,for cases with no outliers, we demonstrate shorter execution timesthan existing optimal algorithms.

Robust curb and ramp detection for safe parking using the Canesta TOF camera

Abstract: Range sensors for assisted backup and parking have potential for saving human lives and for facilitating parking in challenging situations. However, important features such as curbs and ramps are difficult to detect using ultrasonic or microwave sensors. TOF imaging range sensors may be used successfully for this purpose. In this paper we present a study concerning the use of the Canesta TOF camera for recognition of curbs and ramps. Our approach is based on the detection of individual planar patches using CC-RANSAC, a modified version of the classic RANSAC robust regression algorithm. Whereas RANSAC uses the whole set of inliers to evaluate the fitness of a candidate plane, CC-RANSAC only considers the largest connected components of inliers. We provide experimental evidence that CC-RANSAC provides a more accurate estimation of the dominant plane than RANSAC with a smaller number of iterations.

Update Policy of Dense Maps: Efficient Algorithms and Sparse Representation

Abstract: Providing a robot with a fully detailed map is one appealing key for the Simultaneous Localisation and Mapping (SLAM) problem. It gives the robot a lot of hints to solve either the data association or the localisation problem itself. The more details are in the map, the more chances are that different places may appear differently, solving ambiguities. The more landmarks are used, the more accurate are the algorithms that solve the localisation problem since in a least square sense an approximation of the solution is more precise. Last, it helps a lot in the presence of a few dynamic objects because these moving parts of the environment remain marginal in the amount of data used to model the map and can thus be filtered out. For instance, the moving objects can be detected or cancelled in the localisation procedure by robust techniques using Monte-Carlo algorithms [6] or RANSAC [4].

Image Mosaic Based on SIFT

Abstract: The traditional feature-based algorithm was found to be sensitive to rotations and scales. In this paper, an automatic image mosaic technique based on SIFT (Scale Invariant Feature Transform) was proposed by using the rotation and scale invariant property of SIFT. Keypoints are first extracted by searching over all scales and image locations, then the descriptors defined on the keypoint neighborhood are computed, through to compare the Euclidean distance of their descriptors to extract the initial feature points pair, then eliminate spurious feature points pair by applying RANSAC, finally transform the input image with the correct mapping model for image fusion and complete image stitching. Experimental results demonstrate the proposed algorithm is robust to translation, rotation, noise and scaling.

A graph-based approach to symmetry detection

Abstract: Symmetry detection aims at discovering redundancy in the form of reoccurring structures in geometric objects. In this paper, we present a new symmetry detection algorithm for geometry represented as point clouds that is based on analyzing a graph of surface features. We combine a general feature detection scheme with a RANSAC-based randomized subgraph searching algorithm in order to reliably detect reoccurring patterns of locally unique structures. A subsequent segmentation step based on a simultaneous region growing variant of the ICP algorithm is applied to verify that the actual point cloud data supports the pattern detected in the feature graphs. We apply our algorithm to synthetic and real-world 3D scanner data sets, demonstrating robust symmetry detection results in the presence of scanning artifacts and noise. The modular and flexible nature of the graph-based detection scheme allows for easy generalizations of the algorithm, which we demonstrate by applying the same technique to other data modalities such as images or triangle meshes.

Image Matching for Geomorphic Measurement Based on SIFT and RANSAC Methods

Abstract: To effectively realize the image feature matching for geomorphic reverse measurement and rebuilding, a new matching scheme is presented, where the SIFT method are adopted to implement initial geomorphic image matching by going through five stages: scale-space construction, scale-space extrema detection, orientation assignment, keypoint descriptor and feature vector matching. Then, in order to eliminate the wrong matching features existing in the initial matching process, RANSAC algorithm is applied. The experimental results show that this algorithm can effectively improve the accuracy and efficiency of geomorphic image matching.

Failure Detection and Exclusion via Range Consensus

Abstract: With the rise of enhanced GNSS services over the next decade (i.e. the modernized GPS, Galileo, GLONASS, and Compass constellations), the number of ranging sources (satellites) available for a positioning will significantly increase to more than double the current value. One can no longer assume that the probability of failure for more than one satellite within a certain timeframe is negligible. To ensure that satellite failures are detected at the receiver is of high importance for the integrity of the satellite navigation system. With a large number of satellites, it will be possible to reduce multipath effects by excluding satellites with a pseudorange bias above a certain threshold. The scope of this work is the development of an algorithm that is capable of detecting and identifying all such satellites with a bias higher than a given threshold. The Multiple Hypothesis Solution Separation (MHSS) RAIM Algorithm (Ene, 2007; Pervan, et al., 1998) is one of the existing approaches to identify faulty satellites by calculating the Vertical Protection Level (VPL) for subsets of the constellation that omit one or more satellites. With the aid of the subset showing the best (or minimum) VPL, one can expect to detect satellite faults if both the ranging error and its influence on the position solution are significant enough. At the same time, there are geometries and range error distributions where a different satellite, other than the faulty one, can be excluded to minimize the VPL. Nevertheless, with multiple constellations present, one might want to exclude the failed satellite, even if this does not always result in the minimum VPL value, as long as the protection level stays below the Vertical Alert Limit (VAL). The Range Consensus (RANCO) algorithm, which is developed in this work, calculates a position solution based on four satellites and compares this estimate with the pseudoranges of all the satellites that did not contribute to this solution. The residuals of this comparison are then used as a measure of statistical consensus. The satellites that have a higher estimated range error than a certain threshold are identified as outliers, as their range measurements disagree with the expected pseudoranges by a significant amount given the position estimate. All subsets of four satellites that have an acceptable geometric conditioning with respect to orthogonality will be considered. Hence, the chances are very high that a subset of four satellites that is consistent with all the other “healthy” satellites will be found. The subset with the most inliers is consequently utilized for identification of the outliers in the combined constellation. This approach allows one to identify as many outliers as the number of satellites in view minus four satellites for the estimation, and minus at least one additional satellite, that confirms this estimation. As long as more than four plus at least one satellites in view are consistent with respect to the pseudoranges, one can reliably exclude the ones that have a bias higher than the threshold. This approach is similar to the Random Sample Consensus Algorithm (RANSAC), which is applied for computer vision tasks (Fischler, et al., 1981), as well as previous Range Comparison RAIM algorithms (Lee, 1986). The minimum necessary bias in the pseudorange that allows RANCO to separate between outliers and inliers is smaller than six times the variance of the expected error. However, it can be made even smaller with a second variant of the algorithm proposed in this work, called Suggestion Range Consensus (S-RANCO). In S-RANCO, the number of times when a satellite is not an inlier of a set of four different satellites is computed. This approach allows the identification of a possibly faulty satellite even when only lower ranging biases are introduced as an effect of the fault. The batch of satellite subsets to be examined is preselected by a very fast algorithm that considers the alignment of the normal vectors between the receiver and the satellite (first 3 columns of the geometry matrix). Concerning the computational complexity, only 4 by 4 matrices are being inverted as part of both algorithms. With the reliable detection and identification of multiple satellites producing very low ranging biases, the resulting information will also be very useful for existing RAIM Fault Detection and Elimination (FDE) algorithms (Ene, et al., 2007; Walter, et al., 1995).

On-board robust vehicle detection and tracking using adaptive quality evaluation

Abstract: This paper presents a robust method for real-time vehicle detection and tracking in dynamic traffic environments. The proposed strategy aims to find a trade-off between the robustness shown by time-uncorrelated detection techniques and the speed-up obtained with tracking algorithms. It combines both advantages by continuously evaluating the quality of the tracking results along time and triggering new detections to restart the tracking process when quality falls behind a certain quality requirement. Robustness is also ensured within the tracking algorithm with an outlier rejection stage and the use of stochastic filtering. Several sequences from real traffic situations have been tested, obtaining highly accurate multiple vehicle detections.

Analysing Driver's Attention Level using Computer Vision

Abstract: This paper presents a system for evaluating the attention level of a driver using computer vision. The system detects head movements, facial expressions and the presence of visual cues that are known to reflect the user's level of alertness. The fusion of these data allows our system to detect both aspects of inattention (drowsiness and distraction), improving the reliability of the monitoring over previous approaches mainly based on detecting only one (drowsiness). Head movements are estimated by robustly tracking a 3D face model with RANSAC and POSIT methods. The 3D model is automatically initialized. Facial expressions are recognized with a model-based method, where different expressions are represented by a set of samples in a low-dimensional manifold in the space of deformations. The system is able to work with different drivers without specific training. The approach has been tested on video sequences recorded in a driving simulator and in real driving situations. The methods are computationally efficient and the system is able to run in real-time.

Experimental Evaluation of Relative Pose Estimation Algorithms.

Abstract: We give an extensive experimental comparison of four popular relative pose (epipolar geometry) estimation algorithms: the eight, seven, six and five point algorithms. We focus on the practically important case that only a single solution may be returned by automatically selecting one of the solution candidates, and investigate the choice of error measure for the selection. We show that the five point algorithm gives very good results with automatic selection. As sometimes the eight point algorithm is better, we propose a combination algorithm which selects from the solutions of both algorithms and thus combines their strengths. We further investigate the behavior in the presence of outliers by using adaptive RANSAC, and give practical recommendations for the choice of the RANSAC parameters. Finally, we verify the simulation results on real data.

3D Visual Odometry for Road Vehicles

Abstract: This paper describes a method for estimating the vehicle global position in a network of roads by means of visual odometry. To do so, the ego-motion of the vehicle relative to the road is computed using a stereo-vision system mounted next to the rear view mirror of the car. Feature points are matched between pairs of frames and linked into 3D trajectories. Vehicle motion is estimated using the non-linear, photogrametric approach based on RANSAC. This iterative technique enables the formulation of a robust method that can ignore large numbers of outliers as encountered in real traffic scenes. The resulting method is defined as visual odometry and can be used in conjunction with other sensors, such as GPS, to produce accurate estimates of the vehicle global position. The obvious application of the method is to provide on-board driver assistance in navigation tasks, or to provide a means for autonomously navigating a vehicle. The method has been tested in real traffic conditions without using prior knowledge about the scene nor the vehicle motion. We provide examples of estimated vehicle trajectories using the proposed method and discuss the key issues for further improvement.

Application of Image SIFT Features to the Context of CBIR

Abstract: This paper is mainly concerned with the application of a kind of distinctive local invariant feature i.e. Lowe's SIFT feature for the purpose of CBIR, instead of the usually used global feature and local statistical feature based on image segmentation. In our CBIR system, the visual contents of the query image and the database images are extracted and described by the 128-dimensional SIFT feature vectors. The KD-tree with the best bin first(BBF), an approximate nearest neighbors(ANN) search algorithm, is used to index and match those SIFT features. As our contribution, a modified voting scheme called nearest neighbor distance ratio scoring (NNDRS) is put forward to calculate the aggregate scores of the corresponding candidate images in the database respectively. By sorting the database images according to their aggregate scores in descending order, the top few similar images are shown to users as the retrieval results. Additionally, RANSAC can be adopted as a geometry verification method to re-check the results and remove the false matches. Experiments show that our approach can obtain high recall and high precision in the context of CBIR on the famous image databases ZuBud.

Robust 3D Line Extraction from Stereo Point Clouds

Abstract: The paper describes a robust method to extract 3D lines from stereo point clouds. This method combines 2D image information with 3D point clouds from a stereo camera. 2D lines are first extracted from the image in the stereo pair, followed by 3D line regression from the back-projected 3D point set of the images points in the detected 2D lines. In this paper, random sample consensus (RANSAC) is used to estimate 3D line from the 3D point set, the Mahalanobis distance from each 3D point to the 3D line is derived, and the statistically motivated distance measure is used to compute the support for the detected 3D line. Experimental results on real environment with high level of clutter, occlusion, and noise demonstrate the robustness of the algorithm.

Automatic Registration Based on Improved SIFT for Medical Microscopic Sequence Images

Abstract: Automatic registration in microscopic image sequence is a classical problem, which has not been solved well so far. According to the features of medical microscopic image sequence, the SIFT feature detection method of microscopic image registration is introduced. As large dimension of the traditional SIFT descriptor and its complex algorithm, an improved algorithm of the SIFT is presented, which can reduce the dimension. And a two-way matching algorithm is adopted to eliminate repeated matching points. Random Sampling Consensus algorithm (RANSAC) is applied for removal the wrong matching points to improve the accuracy of matching further. Compared with traditional registration algorithm, the results show that the improved SIFT algorithm has increased both in time-saving and complexity-reducing.

Accurate and robust ego-motion estimation using expectation maximization

Abstract: A novel robust visual-odometry technique, called EM-SE(3) is presented and compared against using the random sample consensus (RANSAC) for ego-motion estimation. In this contribution, stereo-vision is used to generate a number of minimal-set motion hypothesis. By using EM-SE(3), which involves expectation maximization on a local linearization of the rigid-body motion group SE(3), a distinction can be made between inlier and outlier motion hypothesis. At the same time a robust mean motion as well as its associated uncertainty can be computed on the selected inlier motion hypothesis. The data-sets used for evaluation consist of synthetic and large real-world urban scenes, including several independently moving objects. Using these data-sets, it will be shown that EM-SE(3) is both more accurate and more efficient than RANSAC.