F. Meyer and S. Beucher, "Morphological segmentation, " J. Vis. Commun. Image Represent., vol. 1, pp. 21--46, 1990.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....obtain a background image aligned in time with the current input frame. Then a foreground mask is extracted for each input video frame. This mask is obtained by comparing the background image with the current video frame. The contours of foreground regions are obtained with a watershed algorithm [11, 12]. Finally, the two last blocks track the resulting foreground masks in time, combine them and model the keyregions for the entire shot. 4.2. Foreground mask extraction The mosaic alignment uses the previously computed mosaic image and the warping parameters ##### to align in time the mosaic ....

F. Meyer and S. Beucher, "Morphological segmentation," Journal of Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--46, September 1990.

....(5) Both and are exploited in the next section to find the object contour . IV. CONTOUR DETECTION We now move to the most important step in the tracking process: the detection of the object contour . For contour detection, we employ the watershed algorithm from mathematical morphology [17] [20], using the two regions and as markers. The advantage of using the watershed algorithm on a set of two markers is that the resulting segmentation always contains only two regions and the resulting contour is always closed and simple. Furthermore, the contour is positioned at the most significant ....

....algorithm, it is helpful to consider how the watershed algorithm is applied in case only one edge indicator function is used. A. Definition of the Watershed This subsection gives the definition of the watershed algorithm applied for one relief function using two markers and . According to [20], in case of the watershed with imposed markers, the original relief is reconstructed. This reconstruction makes the markers sole regional minima of the reconstructed relief. Let be the reconstructed relief. The function can be defined via the recursive conditional erosion as proposed in [20] can ....

[Article contains additional citation context not shown here]

F. Meyer and S. Beucher, "Morphological segmentation," J. Vis. Commun. Image Represent., vol. 1, no. 1, pp. 21--46, Sept. 1990.

....the size of for curve fitting below. Experimentation shows should have a minimum size of , gives . Peak extraction (2) Since is not known, we cannot use the peak extraction (1) method. Instead, we use a watershed segmentation algorithm inspired from image processing [8]. Curve fitting: We have analytical equations for the DD curves. We also have the coordinates c,I o of the peaks just extracted. We can thus perform a LS fit of a parametric DD curve to the peaks. The result is the LS estimate of . The statistics that is minimized is t R i ....

F. Meyer and S. Beucher, "Morphological segmentation," Journal of Visual Communication and Image Representation, vol. 1, no. 1, pp. 121--146, 1990.

....causes many local minima, as illustrated in Fig. 1. Computed in 2D or in 3D, each local minimum will be the starting point of a basin and thus will cause oversegmentation. Various solutions can be used to improve this situation. The most efficient one is the use of makers, as firstly proposed in [23]. The first 3D watershed algorithm using markers and hierarchical queues was proposed by Gratin [9] This leads to efficient 3D segmentation of medical images. But the use of markers often requires various pre processing steps, such as filtering, contrast enhancement or definition of inner and ....

F. Meyer, S. Beucher, Morphological segmentation, J. Comm. Image Represent. 1 (1) (September 1990) 21-46.

....level set model and present an implementation performing at ms for an explicit image. 1. INTRODUCTION Robust and efficient segmentation algorithms on digital images are a key task in many computer vision applications. Morphological techniques such as the watershed transform and its successors [1, 2, 3] based on the simulated rainfall on the topographically interpreted image, and propagation techniques such as the active contour models [4, 5, 6, 7, 8] based on a controlled curve or surface evolution, have been presented recently. Especially the latter have proven very flexible in incorporating ....

F. Meyer and S. Beucher, "Morphological segmentation, " J. Vis. Commun. Image Represent., vol. 1, pp. 21--46, 1990.

....and minimum description length (MDL) criteria [19] It is found to be computationally costly to compute the globally optimal solutions. The last group of segmentation methods is hybrid techniques which combine boundary and region criteria and may result in better segmentation. Meyer and Beucher [21] proposed a morphological segmentation approach based on watershed. Besl and Jain [3] presented a segmentation method through variable order surface fitting. Pavlidis and Liow [25] designed a method to integrate region growing and edge detection to produce smoother contours. Zhu and Yuille [31] ....

F. Meyer and S. Beucher, "Morphological Segmentation," J. Vision Comm. and Image Representation, vol. 1, pp. 21-46, 1990.

....abilities can become a major analysis tool in the multimedia oriented standards MPEG 4 and MPEG 7. Several techniques and algorithms have been proposed in the literature for image segmentation. Some color oriented methods have been recently proposed based on the morphological watershed [5] or by using split and merge techniques [6] However, an intrinsic property of video objects is that they usually consist of regions of totally different color characteristics and consequently the main problem of any color oriented segmentation scheme is that it oversegments an object into ....

F. Meyer and S. Beucher, "Morphological Segmentation," Journal of Visual Communication on Image Representation, Vol. 1, No. 1, pp.21-46, September 1990.

....merges, at the same time, partial regions, produced by the local labeling, and nonmarked regions to marked basins. Both modules are designed with great concurrency, locality, and reduced software engineering cost, emerging into a scalable algorithm. 1 Introduction Watershed transformation [2, 10, 11] is an effective tool for morphological image segmentation. The sequential algorithm starts by detecting and labeling initial seeds, e.g. regional minima of the gradient (connected plateaus of constant altitude which do not have neighboring pixels of lower graylevel) Starting from minima, ordered ....

F. Meyer and S. Beucher, "Morphological Segmentation," in Journal of Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--46, September 1990. 16

....iteration routine. Other recent models of feature binding in contour processing have only considered gure ground segmentation ( Yen and Finkel, 1998] Li, 1999] Segmentation algorithms in related computer vision applications are for example based on (i) ooding schemes in gradient mountains [Meyer and Beucher, 1990], ii) wave propagation [Hanahara and Hiyane, 1990] iii) boundary tracing [Galbraith et al. 1991] or (iv) backtracking search through combinations of line segments by calculating a saliency measure [Jacobs, 1996] In our present study the considered objects are lymphocytes during a migration in ....

Meyer, F. and Beucher, S. (1990). Morphological segmentation. J. Visual Communications and Image Representation, 1(1):21-46.

....to propagate. At each pixel, the flow points towards the estimated direction of the boundary pixel. A boundary location can then be characterized by flows in opposing directions toward it. Another popular approach to spatial segmentation uses the watershed algorithm as the basic decision tool [38, 54]. The watershed technique is a region growing algorithm that analyzes the image as a topographic surface. It detects the minima of the gradient of the gray level image and grows these minima according to the gradient values much like a flooding process. Points of contact between the propagation ....

F. Meyer and S. Beucher. Morphological segmentation. J. Visual Commun. Image Represent., 1(1):21--46, Sept. 1990.

....the image. These lters have become very popular in image processing because, on an experimental basis, they have been claimed to simplify the image while preserving contours. This property has made them very attractive for a large number of applications, such as noise cancellation or segmentation [17, 33]. More recently, they have become the basis of a morphological approach to image and video compression [23, 24, 22, 6] Di erent classes of connected operators have been studied by Meyer [15, 16] Serra [29] or Heijmans [8] see also references therein) In this article, we study the theoretical ....

Meyer, F. and S. Beucher: 1990, `Morphological Segmentation'. Journal of Visual Communication and Image Representation 1(1), 21-46.

....in this paper belongs to the category of hybrid techniques, since it results from the integration of edge and region based techniques through the morphological watershed transform. Many morphological segmentation approaches using the watershed transform have been proposed in the literature [30], 31] Watersheds have also been used in multiresolution methods for producing resolution hierarchies of image ridges and valleys [3] 32] Although these methods were successful in segmenting certain classes of images, they require significant interactive user guidance or accurate prior ....

....representations required by higher level analysis, namely, onepixel wide closed curves. A different approach to image segmentation, which overcomes the problem of disconnected contours and false edges, is the application of the morphological watershed transform to the gradient magnitude image [30], 31] This approach allows the generation of an initial image partition into regions and, consequently, region based techniques can be used in order to produce closed, one pixel wide contours surfaces. In the following, we briefly describe the fast watershed detection algorithm proposed by ....

[Article contains additional citation context not shown here]

F. Meyer and S. Beucher, "Morphological segmentation," J. Vis. Commun. Image Represent., vol. 1, pp. 21--46, Sept. 1990.

....topographic map of the image. These lters have become very popular because, on an experimental basis, they have been claimed to simplify the image while preserving contours. This property has made them very attractive for a large number of applications such as noise cancellation or segmentation ([21], 46] More recently, they have become the basis of a morphological approach to image and video compression (see [40] and references therein, and more recently [25] 12 4 Structure of the simpli ed topographic map Assuming that the connected components of the upper and lower level sets are ....

F. Meyer and S. Beucher, Morphological Segmentation. J. Visual Commun. Image Representation, 1 (1990), 21-46.

....not least it should be fast. Classical gradient based edge detectors such as a Sobel operator or Kirsch masks are not sucient for this task, as they do not give closed contours. This also holds for more sophisticated variants such as the Canny edge detector [31] We found a watershed technique [32, 33] based on the squared gradient magnitude very useful for these purposes. Such a technique regards an image as a landscape where the intensity values correspond to the elevation. Areas where a rain drop would drain to the same minimum are denoted as catchment basins, and the lines separating ....

....algorithms often create too many segments. Although this oversegmentation is less dominant in the PDE regularized image than in the original one, it may still lead to problems. Numerous ways have been proposed in order to deal with the oversegmentation problem, for instance by using markers [33, 39], region merging [40, 41, 10] or scale space techniques [13, 17, 14] In our case we shall see that a simple region merging strategy is adequate. In such a step, adjacent regions are merged if their contrast di erence is below a speci ed threshold. This contrast parameter can be related to the ....

F. Meyer, S. Beucher, Morphological segmentation, J. Vis. Comm. Image Repr., Vol. 1, 21-46, 1990.

....if desired. Other recent models of feature binding in contour processing have only considered gure ground segmentation ( Yen and Finkel, 1998] Li, 1999] Segmentation algorithms in related computer vision applications are for example based on (i) ooding schemes in gradient mountains [Meyer and Beucher, 1990], ii) wave propagation [Hanahara and Hiyane, 1990] iii) boundary tracing [Galbraith et al. 1991] or (iv) backtracking search through combinations of line segments by calculating a saliency measure [Jacobs, 1996] In our present study the considered objects are lymphocytes during a migration in ....

Meyer, F. and Beucher, S. (1990). Morphological segmentation. J. Visual Communications and Image Representation, 1(1):21-46.

....are used as focus points for guiding the system to the cells in the image. In the focused image patches a recurrent neural network is used to extract the cell contour. Segmentation algorithms in related computer vision applications are usually based on (i) flooding schemes in gradient mountains [Meyer and Beucher, 1990], ii) wave propagation [Hanahara and Hiyane, 1990] iii) boundary tracing [Galbraith et al. 1991] or (iv) backtracking search through combinations of line segments by calculating a saliency measure [Jacobs, 1996] In our present study the considered objects are lymphocytes in tissue, so their ....

Meyer, F. and Beucher, S. (1990). Morphological segmentation. J. Visual Communications and Image Representation, 1(1):21--46.

....various artistic effects, as shown in Fig. 2. There we also see that the surface evolution GWDT gives a smoother halftoning of the image than the GWDTs based on chamfer metrics. 1.2. 4 Watershed Segmentation via Eikonal A powerful morphological approach to image segmentation is the watershed [14, 25] which transforms an image f(x, y) to the crest lines separating adjacent catchment basins that surround regional minima or other marker sets of feature points. In [13, 15] it has been established that (in the continuous domain and assuming that the image is smooth and has isolated critical ....

....the gradient magnitude of the filtered image, and changing (via morphological reconstruction) the homotopy of the gradient image so that its only minima occur at the markers. The segmentation is done on the final outcome f of the above processing. In the standard digital watershed algorithm [25, 14], the flooding at each level is achieved by a planar distance propagation that uses the chess board metric. This kind of distance propagation is non isotropic and could give wrong results, particularly for images with large plateaus, as we found experimentally. Eikonal segmentation using ....

F. Meyer and S. Beucher, "Morphological Segmentation ", J. Visual Commun. Image Representation, 1(1):21--45, 1990.

....map of the image. These filters have become very popular because, on an experimental basis, they have been claimed to simplify the image while preserving contours. This property has made them very attractive for a large number of applications such as noise cancellation [64, 65] or segmentation [47, 66]. More recently, they have become the basis of a morphological approach to image and video compression (see [59] and references therein, and more recently [27] Application to connected operators As an application of the theory of M connected components for sets of finite perimeter developed ....

F. Meyer and S. Beucher, Morphological Segmentation. J. Visual Commun. Image Representation, 1 (1990), 21--46.

....FEATURE EXTRACTION As part of a large project in classifying cells on cervical smears, we are interesting in characterising the visual texture of stained nuclei. Such texture is one of the cell classification cues used by trained cytologists and many features have been reported in the literature [11, 3, 12, 4]. In this paper we wish to extract classification features from the watershed based nuclear regions. Since our features can be made scale dependant, in continuing work we are investigating approaches such as classification in scale space [13] however the results reported here were obtained at a ....

F. Meyer and S. Beucher, "Morphological segmentation," J. Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--46, 1990.

....appendix. Xi 3. The Extraction of Texture Features As part of a large project in classifying cells on cervical smears, we are interesting in characterising the visual texture of stained nucleii. Such texture has been reported as being useful in discriminating between normal and abnormal cells [12, 13]. This work is still in progress, and the results are yet to be reported, however, a typical set of multiscale images showing gradient watershed lines (Fig. 1) serves to illustrate the application of the technique developed in the current paper. In the cell classification project, several ....

F. Meyer and S. Beucher, "Morphological segmentation," Journal of Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--46, 1990.

....) is the average graylevel of W . As part of a large project in classifying cells from cervical smears, we are interested in characterising the visual texture of Papanicolaou stained nuclei. Such texture has been reported as being useful in discriminating between normal and abnormal cells [36] [37] and a multiscale approach to texture analysis has been used [38] A typical set of multiscale images of a cell nucleus showing gradient watershed lines (Fig. 3) serves to illustrate the application of the proposed technique. In the cell classificFig. 3 near here. Fig. 3. ation project, the above ....

F. Meyer and S. Beucher, "Morphological segmentation", Journal of Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--46, 1990.

....are a large number of simple, low level segmentation algorithms available (Haralick and Shapiro (1985) Jain, Kasturi and Schunck (1995) Chapter 3) These algorithms provide an initial partitioning of an image. For concreteness, in this paper we will consider the watershed algorithm (Serra (1982) Meyer and Beucher (1990), Vincent and Soille (1991) which automatically provides a segmentation according to definition (2) above. Figure 3 depicts the results of a watershed segmentation of our example mammogram. In fact, as the field under consideration is actually the coefficient of variation field shown in Figure ....

Meyer, F., and Beucher, S. (1990), "Morphological Segmentation," Journal of Visual Communication and Image Representation, 1, 21-46.

....FEATURE EXTRACTION As part of a large project in classifying cells on cervical smears, we are interesting in characterising the visual texture of stained nuclei. Such texture is one of the cell classification cues used by trained cytologists and many features have been reported in the literature [11, 3, 12, 4]. In this paper we wish to extract classification features from the watershed based nuclear regions. Since our features can be made scale dependent, in continuing work we are investigating approaches such as classification in scale space [13] however the results reported here were obtained at a ....

F. Meyer and S. Beucher, "Morphological segmentation," J. Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--46, 1990.

No context found.

F. Meyer and S. Beucher. Morphological segmentation. 1(1):2146, Sept. 1990.

....these borders into the sensomotoric pathways within the central lobe and or the basal ganglia. 2 Review of related work In computer vision literature various methods dealing with segmentation and feature extraction are discussed. The well known technique of the morphological watershed transform [13] creates a tesselation of the image domain# in several small regions by considering the image values as intensity niveaus in a topographical landscape. By simulating rainfall, the domain is grouped in catchment basins, regions in which the water drains from all points to the same local intensity ....

F. Meyer and S. Beucher. Morphological segmentation. J. Vis. Commun. Image Represent., 1:21--46, 1990.

....the problem of this constantly present error. In [1] we have proposed a complete system which eliminates the error and which, based on its position deduces the depth ordering of the objects. This technique consists of the following stages: still image segmentation using the watershed algorithm [4], motion estimation via a gradient technique (cf. equation (1) combining of the motion measurements in the regions through adjustment of a parametric model by robust regression, region merging and depth ordering. In this paper we concentrate on the depth ordering. Figure 10(a) shows another ....

F. Meyer and S. Beucher. Morphological segmentation. J. Visual Commun. Image Repres., 1(1):2145, 1990.

....and divide lines to segmentation problems [3] They used it to segment 1 images of bubbles and SEM metallographic pictures. Unfortunately, this transformation very often leads to an over segmentation of the image. To overcome this problem, a strategy has been proposed by Meyer and Beucher [7]. This strategy is called marker controlled segmentation. This approach is based on the idea that machine vision systems often roughly know from other sources the location of the objects to be segmented. This approach is applied as follows : first, we define the properties which will be used ....

Meyer F. and Beucher S., "Morphological Segmentation", Journal of Visual Communication and Image representation, Vol. 1, n 1, pp. 21-46, Sept. 1990, Academic Press.

....are good enough when the contrast of the images is sufficient. These approach needs, however, the setting of many parameters, which compromise the robustness of the process. The second approach uses exclusively morphological segmentation tools for based on a transformation called watersheds [3]. Image segmentation by watersheds is a technique of segmentation based on the marking of the objects to be extracted. This marking is the fundamental step of the process. We will discuss more about these tools in the next sections. 3.1 Image Segmentation. An overview A grey tone image can be ....

F. MEYER, S. BEUCHER, Morphological Segmentation, Journal of Visual Computing, in press

No context found.

F. Meyer and S. Beucher, "Morphological segmentation, " J. Vis. Commun. Image Represent., vol. 1, pp. 21--46, 1990.

No context found.

F. Meyer and S. Beucher. Morphological segmentation. J. Vis. Commun. Image Represent., 1:21--46, 1990.

No context found.

F. Meyer and S. Beucher, "Morphological segmentation, " Journal of Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--46, 1990.

No context found.

F. Meyer and S. Beucher, "Morphological Segmentation," Journal of Visual Communication on Image Representation, Vol. 1, No. 1, pp.21-46, September 1990.

No context found.

F. Meyer and S. Beucher, "Morphological segmentation," J. Visual Commun. Image Representation, vol. 1, no. 1, pp. 21--46, Sept. 1990.

No context found.

F. Meyer and S. Beucher, "Morphological Segmentation," Journal of Visual Communication on Image Representation, Vol. 1, No. 1, pp.21-46, September 1990.

No context found.

Meyer and S. Beucher, "Morphological Segmentation," Journal of Visual Communication on Image Representation, Vol. 1, No. 1, pp.21-46, September 1990.

No context found.

Meyer and S. Beucher, "Morphological Segmentation," Journal of Visual Communication on Image Representation, Vol. 1, No. 1, pp.21-46, September 1990.

No context found.

Meyer and S. Beucher, "Morphological segmentation," J. Visual Commun. Image Representation, vol. 1, no. 1, pp. 21--46, Sept. 1990.

No context found.

F. Meyer and S. Beucher, "Morphological Segmentation," Journal of Visual Communication on Image Representation, vol. 1, no. 1, pp.21-46, September 1990.

No context found.

F. Meyer and S. Beucher, "Morphological segmentation," J. Visual Commun. Image Representation, vol. 1, pp. 21--46, Sept. 1990.

No context found.

F. Meyer and S. Beucher, "Morphological Segmentation," Visual Comm. Image. Representaion, vol. 1, no. 1, pp. 21-46 , Sept. 1990.

No context found.

F. Meyer and S. Beucher, "Morphological segmentation," Journal of Visual Comm. and Image Rep., vol. 1, no. 1, pp. 21--46, Sept. 1990.

No context found.

Meyer, F. and Beucher, S. (1990). Morphological segmentation. J. Visual Communications and Image Representation, 1(1):21--46.

No context found.

F. Meyer and S. Beucher, "Morphological segmentation", Journal of Visual Com. and Im. Proc., Vol. 1, No 1, pp. 21-46 (1990).

No context found.

F. Meyer and S. Beucher, "Morphological Segmentation, " Journal of Visual Communication and Image Representation, vol. 1, no. 1, pp. 21--45, 1990. 56

No context found.

F. Meyer, S. Beucher, Morphological segmentation, JVCIR 1 (1) (1990) 21--46.

No context found.

F. Meyer and S. Beucher, "Morphological segmentation," Journal of Visual Comm. and Image Rep., vol. 1, no. 1, pp. 21--46, Sept. 1990.

No context found.

Meyer, F. and S. Beucher, Morphological Segmentation. J. Visual Comm. Image Rep., 1990. 1(1): p. 21-46.

No context found.

F. Meyer and S. Beucher, "Morphological Segmentation," J. Visual Comm. and Image Representation, vol. 1, no. 1, pp. 21-46, Sept. 1990.

No context found.

F. Meyer and S. Beucher. Morphological segmentation. J. Visual Commun. Image Repres., 1(1):21--45, 1990. 833

No context found.

F. Meyer and S. Beucher, "Morphological segmentation," Journal of Visual Communication and Image Representation 1(1), pp. 21--46, 1990.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC