A new approach toward target representation and localization, the central component in visual tracking of non-rigid objects, is proposed. The feature histogram based target representations are regularized by spatial masking with an isotropic kernel. The masking induces spatially-smooth similarity functions suitable for gradient-based optimization, hence, the target localization problem can be formulated using the basin of attraction of the local maxima. We employ a metric derived from the Bhattacharyya coefficient as similarity measure, and use the mean shift procedure to perform the optimization. In the presented tracking examples the new method successfully coped with camera motion, partial occlusions, clutter, and target scale variations. Integration with motion filters and data association techniques is also discussed. We describe only few of the potential applications: exploitation of background information, Kalman tracking using motion models, and face tracking. Keywords: non-rigid object tracking; target localization and representation; spatially-smooth similarity function; Bhattacharyya coefficient; face tracking. 1

This paper aims to present a review of recent as well as classic image registration methods. Image registration is the process of overlaying images (two or more) of the same scene taken at different times, from different viewpoints, and/or by different sensors. The registration geometrically align two images (the reference and sensed images). The reviewed approaches are classified according to their nature (areabased and feature-based) and according to four basic steps of image registration procedure: feature detection, feature matching, mapping function design, and image transformation and resampling. Main contributions, advantages, and drawbacks of the methods are mentioned in the paper. Problematic issues of image registration and outlook for the future research are discussed too. The major goal of the paper is to provide a comprehensive reference source for the researchers involved in image registration, regardless of particular application areas. q 2003 Elsevier B.V. All rights reserved.

In this paper, we first propose a new subdivision of the image information axis uis for the classification of nonrigid registration algorithms. Namely, we introdu) the notion of iconic featuy based (IFB) algorithms, which lie between geometrical and standard intensitybased algorithms fortheyuM both anintensitysimilaritymeasu and a geometrical distance. Then we present a new registration energyfor IFB registration that generalizes some of the existing techniquML We compareou algorithm with other registration approaches, and show the advantages of this energy. Besides, we also present a fasttechniqu for thecompukUy)- of local statistics between images, which tuchou to beuy-UM on pairs of images having a complex, nonstationaryrelationship between their intensities, as well as an hybridreguSkq-y)-qL scheme mixing elastic and fluy components. The potential of the algorithm is finallydemonstrated on a clinical application, namelydeep brainstimuMUq-y of a Parkinsonian patient. Registration of pre- and immediate postoperative MR images allow toqu--MSy)WS range of the deformationdu topneuU3y)W3Mflover the entire brain,thu yielding tomeasuMy)W3 of the deformation aroun the preoperatively computed stereotactic targets.

This paper presents an original method for three-dimensional elastic registration of multimodal images. We propose to make use of a scheme that iterates between correcting for intensity differences between images and performing standard monomodal registration. The core of our contribution resides in providing a method that finds the transformation that maps the intensities of one image to those of another. It makes the assumption that there are at most two functional dependences between the intensities of structures present in the images to register, and relies on robust estimation techniques to evaluate these functions. We provide results showing successful registration between several imaging modalities involving segmentations, T1 magnetic resonance (MR), T2 MR, proton density (PD) MR and computed tomography (CT). We also argue that our intensity modeling may be more appropriate than mutual information (MI) in the context of evaluating highdimensional deformations, as it puts more co...

In this article we merge point feature and intensity-based registration in a single algorithm to tackle the problem of multiple brain registration. Because of the high variabilit of the shape of the cortex across ind vidy ls, there exist geometrical ambiguities in the registration process that an intensit measure alone is unable to solve. This problem can be tackled using anatomical knowledge. First, we automatically segment and label the whole set of the cortical sulci, with a non-parametric approach that enables the capture of their highly variable shape and topology. Then, we develop a registration energy that merges intensity and feature point matching. Its minimization leads to a linear combination of a dense smooth vector field and radial basis functions. We use and process differently the bottom line of the sulci from its upper border, whose localization is even more variable across individuals. We show that the additional sulcal energ improves the registration of the cortical sulci, while still keeping the transformation smooth and one-to-one.

Multi-subject non-rigid registration algorithms using dense deformation fields often encounter cases where the transformation to be estimated has a large spatial variability. In these cases, linear stationary regularization methods are not su#cient. In this paper, we present an algorithm that uses a priori information about the nature of imaged objects in order to adapt the regularization of the deformations. We also present a robustness improvement that gives higher weight to those points in images that contain more information. Finally, a fast parallel implementation using networked personal computers is presented. In order to improve the usability of the parallel software by a clinical user, we have implemented it as a grid service that can be controlled by a graphics workstation embedded in the clinical environment. Results on inter-subject pairs of images show that our method can take into account the large variability of most brain structures. The registration time for images 124 is 5 minutes on 15 standard PCs. A comparison of our non-stationary visco-elastic smoothing versus solely elastic or fluid regularizations shows that our algorithm converges faster towards a more optimal solution in terms of accuracy and transformation regularity.

Abstract—This paper presents an original method for three-dimensional elastic registration of multimodal images. We propose to make use of a scheme that iterates between correcting for intensity differences between images and performing standard monomodal registration. The core of our contribution resides in providing a method that finds the transformation that maps the intensities of one image to those of another. It makes the assumption that there are at most two functional dependencies between the intensities of structures present in the images to register, and relies on robust estimation techniques to evaluate these functions. We provide results showing successful registration between several imaging modalities involving segmentations, T1 magnetic resonance (MR), T2 MR, proton density (PD) MR and computed tomography (CT). We also argue that our intensity modeling may be more appropriate than mutual information (MI) in the context of evaluating high-dimensional deformations, as it puts more constraints on the parameters to be estimated and, thus, permits a better search of the parameter space. Index Terms—Elastic registration, intensity correction, medical imaging, multimodality, robust estimation. I.

While intensity-based similarity measures are increasingly used for medical image registration, they often rely on implicit assumptions regarding the physics of imaging. The motivation of this paper is to determine what are the assumptions corresponding to a number of popular similarity measures in order to better understand their use, and finally help choosing the one which is the most appropriate for a given class of problems. After formalizing registration based on general image acquisition models, we show that the search for an optimal measure can be cast into a maximum likelihood estimation problem. We then derive similarity measures corresponding to di#erent modeling assumptions and retrieve some well-known measures (correlation coe#cient, correlation ratio, mutual information). Finally, we present results of registration between 3D MR and 3D Ultrasound images to illustrate the importance of choosing an appropriate similarity measure.

This manuscript tackles the registration of 2D biological images (histological sections or autoradiographs) to 2D images from the same or di#erent modalities (e.g., histology or MRI). The process of acquiring these images typically induces composite transformations that can be modeled as a number of rigid or a#ne local transformations embedded in an elastic one. We propose a registration approach closely derived from this model. Given a pair of input images, we first compute a dense similarity field between them with a block matching algorithm. A hierarchical clustering algorithm then automatically partitions this field into a number of classes from which we extract independent pairs of sub-images. Finally, the pairs of sub-images are, independently, a#nely registered and a hybrid a#ne/non-linear interpolation scheme is used to compose the output registered image. We investigate the behavior of our approach under a variety of conditions, and discuss examples using real biomedical images, including MRI, histology and cryosection data.

Automatic processing of 3D ultrasound (US) is of great interest for the development of innovative and low-cost computer-assisted surgery tools. In this paper, we present a new image-based technique to rigidly register intra-operative 3D US with pre-operative Magnetic Resonance (MR) data. Automatic registration is achieved by maximization of a similarity measure that generalizes the correlation ratio (CR). This novel similarity measure has been designed to better take into account the nature of US images. A preliminary cross-validation study has been carried out using a number of phantom and clinical data. This indicates that the worst registration errors are of the order of the MR resolution.