Implicit active contours are a very flexible technique in the segmentation of digital images. A novel type of hardware implementation is presented here to approach real time applications. We propose to exploit the high performance of modern graphics cards for numerical computations. Vectors are regarded as images and linear algebraic operations on vectors are realized by the graphics operations of image blending. Thus, the performance benefits from the high memory bandwidth and the economy of command transfers, while the restricted precision does not infect the qualitative behavior of the level set propagation. Here we pick up a first order solver for the basic implicit level set model and present an implementation performing at ms for an explicit image.

An e#cient adaptive multigrid level set method for front propagation purposes in three dimensional medical image processing and segmentation is presented. It is able to deal with non sharp segment boundaries. A flexible, interactive modulation of the front speed depending on various boundary and regularization criteria ensure this goal. E#ciency is due to a graded underlying mesh implicitly defined via error or feature indicating values on the cells of the underlying hexahedral grid. A suitable saturation condition ensures an important regularity condition on the resulting adaptive grid. This simplifies the adaptive fast marching method on the compressed data significantly. As an application the segmentation of glioma is considered. Thus the clinician interactively selects a few parameters describing the speed function and a few seed points referring to a single slice of an MRI data set. Then the automatic process of front propagation generates a family of segments corresponding to the evolution of the front in time, from which the clinician finally selects an appropriate segment covered by the gliom. This selection can be based on a visual evaluation of the propagation on a reference slice using the clinicians expert knowledge. Thus, the overall glioma segmentation turns into an e#cient, nearly real time process with intuitive and usefully restricted user interaction.

Abstract not found

Segmentation is an essential ingredient in a wide range of image processing tasks and a building block of many visualization environments. Many known segmentation techniques su#er from being computationally exhaustive and thus decreasing interactivity, especially when considering volume data sets. Multilevel methods have proved to be a powerful machinery to speed up applications which incorporate some hierarchical structure. So does segmentation when considered on quadtree respectively octree data sets. Here we present a new approach which combines a discrete and a continuous multilevel segmentation model. In figure

The processing of microstructure images dedicated to detection of the borders between material grains is still a difficult task. It is basically caused by a superimposed noise in form of visible scratches and micro inclusions. Thus, the analysis of the microstructure photographs is performed manually in most cases, which is time consuming for numerous set of images. To avoid this problem the approach of automated images processing has been proposed. This approach consists of two parts i.e. edge detection, designed and implemented using Canny Detector method (Ritter & Wilson, 1996) and data filtering, based on Particle Dynamics method (Rauch & Kusiak, 2005a). The results obtained from this approach is in form of new microstructure image with smoothed grain areas and precisely detected grain borders. Such effect allows to optimize further analysis of material structure including e.g. Watershed (Haris et al., 1998) edge fulfilment or statistical calculations of average grain size. The paper presents basic assumptions of proposed approach and details of both algorithms. The results of the analysis of microstructure images using edge detection and filtering algorithms are presented.

This paper deals with mathematical properties of watersheds in weighted graphs linked to region merging methods, as used in image analysis. In a graph, a cleft (or a binary watershed) is a set of vertices that cannot be reduced, by point removal, without changing the number of regions (connected components) of its complement. To obtain a watershed adapted to morphological region merging, it has been shown that one has to use the topological thinnings introduced by M. Couprie and G. Bertrand. Unfortunately, topological thinnings do not always produce thin clefts. Therefore, we introduce a new transformation on vertex weighted graphs, called C-watershed, that always produces a cleft. We present the class of perfect fusion graphs, for which any two neighboring regions can be merged, while preserving all other regions, by removing from the cleft the points adjacent to both. An important theorem of this paper states that, on these graphs, the C-watersheds are topological thinnings and the corresponding divides are thin clefts. We propose a linear-time immersion-like monotone algorithm to compute C-watersheds on perfect fusion graphs, whereas, in general, a linear-time topological thinning algorithm does not exist. Finally, we derive some characterizations of perfect fusion graphs based on thinness properties of both C-watersheds and topological watersheds.

This paper presents a human posture recognition method from a single image. We first segment an image into homogeneous regions and extract curve segments corresponding to human body parts. Each body part is considered as a 2D ribbon. From the smooth curve segments in skin regions, 2D ribbons are extracted and a human body model is constructed. We assign a predefined posture type to the image according to the constructed body model. For the user input query to retrieve images containing human of specific posture, the system convert the query to a body model. The body model is compared to other body models saved in the local storage of target images and images of good matches are retrieved. When a face detection result is available for the given image, it is also used to increase the reliability of body model. For the query human posture, our system retrieves images of the corresponding posture. As another application, the proposed method provides an initial location of a human body to t...

COLORSPACE A hybrid split and merge segmentation method for color images is presented in this work. It combines edge and region information to merge adjacent regions produced in the initial watershed-based segmentation stage. A novel technique is introduced to simplify the Region Adjacency Graph (RAG) structure and speed-up the merging process along with a merging termination criterion for automatic segmentation. The robustness of the proposed method has been experimentally verified and compared to other previously reported merging approaches.

Abstract—This paper proposes a general quadrilateral-based framework for image segmentation, in which quadrilaterals are first constructed from an edge map, where neighboring quadrilaterals with similar features of interest are then merged together to form regions. Under the proposed framework, the quadrilaterals enable the elimination of local variations and unnecessary details for merging from which each segmented region is accurately and completely described by a set of quadrilaterals. To illustrate the effectiveness of the proposed framework, we derived an efficient and high-performance parameterless quadrilateral-based segmentation algorithm from the framework. The proposed algorithm shows that the regions obtained under the framework are segmented into multiple levels of quadrilaterals that accurately represent the regions without severely over or undersegmenting them. When evaluated objectively and subjectively, the proposed algorithm performs better than three other segmentation techniques, namely, seeded region growing, K-means clustering and constrained gravitational clustering, and offers an efficient description of the segmented objects conducive to content-based applications. Index Terms—Approximate methods, object representations, region growing, quadrilateral-based segmentation. 1

In this paper, we propose an image segmentation method based on morphological decomposition and graph-based region merging using contour dynamics. The input image is initially decomposed into a set of primitive homogeneous regions through the morphological watershed transform applied to the image intensity gradient magnitude. This decomposition is represented by a Region Adjacency Graph (RAG) that is input to a hierarchical merging process in which neighboring regions of high similarity are merged. The region similarity criterion is based on the concept of watershed contour dynamics. The robustness of the segmentation to the presence of noise and/or low contrast is improved by a regularization of the contour dynamics. Experimental results on various kinds of synthetic and real images, as well as comparison of the proposed method with other wellknown region merging algorithms are presented.