S.T. Acton and D.P. Mukherjee. Scale space classification using area morphology. IEEE Trans. Image Processing, 9(4):623--635, 2000.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....We post process the connected components to remove any components whose surface area i j is less than some threshold min j (a parameter of the method) to eliminate regions corresponding to noise and artifacts in the original image. This parameter is commonly used in image segmentation [42, 1]. To implement our level set image segmentation based on energy minimization, a four stages method is used. Let L, min j be the input parameters set by the user. 1. Bilevel set construction The first step completes a crude mapping of each image pixel on a given bilevel set. At present, we ....

....labeling. So we consider a size oriented morphological operator acting on sets that consists in keeping all connected components of the output of area larger than a limit min j. This area operator does not introduce new features or edges and boundaries of connected components are preserved [42, 1]. The list of remained connected components then forms the bank C T of admissible T objects T g (P T ) such as min j. The connected components of area lower than min j are a part of the 3. Configuration determination The connected components are then combined during the third step to ....

[Article contains additional citation context not shown here]

S.T. Acton and D.P. Mukherjee. Scale space classification using area morphology. IEEE Trans. Image Processing, 9(4):623--635, 2000.

....a proximity measure that describes the affinity between components within the image level set. This proximity measure helps in region merging for effective image segmentation. The level sets describe a unique representation of the image satisfying properties like causality and edge localization [2]. The proposed segmentation approach based on image level set therefore satisfies these important properties as well. In [2] image segmentation is achieved by sequential processing of area morphological operation followed by clustering using fuzzy c means approach. In this case the component ....

....in region merging for effective image segmentation. The level sets describe a unique representation of the image satisfying properties like causality and edge localization [2] The proposed segmentation approach based on image level set therefore satisfies these important properties as well. In [2] image segmentation is achieved by sequential processing of area morphological operation followed by clustering using fuzzy c means approach. In this case the component merging is achieved as we proceed along the level set stack of any intensity image. Also, we do not need any a priori knowledge ....

S.T. Acton and D.P. Mukherjee, "Scale Space Classification using Area Morphology," IEEE Trans. IP, vol. 9, no. 4, 2000, 623-635.

....components) are not distorted in processing, as with the traditional morphological filters. Moreover, the AOC filter guarantees the removal of connected components beneath the minimum area. Various researchers have shown in the literature the attractive properties of the area morphological filters [2], 9] 11] These include feature causality, Euclidean invariance, and edge preservation through 739 scale. For these reasons, the AOC is optimally suitcd to our scaling task in CBR. Another nonlinear filter that flattens image regions and preserves edges is the median filter [3] In the I D ....

....in scaled texture space The segmentation for CBR is obtained by fuzzy clustering through the AOC scaled texture space. For each position (x, y) in the input image, and each texture layer t, we have vectors l(x, y, t) that are clustered using the multi scale fuzzy c means (FCM) algorithm [2]. The FCM generated segmentations give regions that are homogeneous in terms of texture and are significant in terms of scale. Figure 2 shows some typical clustering results. It can be noticed that due to the staircasing effect of the AOC, a few undesirable segments that do not correspond to ....

S.T. Acton and D.P. Mukherjee, "Scale space classification using area morphology," IEEE Trans. on Image Processing, vol. 9, pp. 623-635, 2000.

....by Level Set Analysis Badrinarayan Raghunathan The Oklahoma Imaging Laboratory School of Electrical and Computer Engineering Oklahoma State University Stillwater, OK 74078 USA Scott T. Acton Deparhnent of Electrical Engineering Thornton Hall University of Virginia Charlottesville, VA 22903 USA Abstract This paper describes an automated image segmentation technique that subdivides regions of homogeneous texture. The method utilizes a level set analysis of scaled Gabor filter responses. Scaling is achieved via an area morphological process. Each scaled, filtered image is ....

....by Level Set Analysis Badrinarayan Raghunathan The Oklahoma Imaging Laboratory School of Electrical and Computer Engineering Oklahoma State University Stillwater, OK 74078 USA Scott T. Acton Deparhnent of Electrical Engineering Thornton Hall University of Virginia Charlottesville, VA 22903 USA Abstract This paper describes an automated image segmentation technique that subdivides regions of homogeneous texture. The method utilizes a level set analysis of scaled Gabor filter responses. Scaling is achieved via an area morphological process. Each scaled, filtered image is ....

[Article contains additional citation context not shown here]

S.T Acton and D.P. Mukherjee, "Scale space classification using area morphology," IEEE Transactions on Image Processing, vol. 9, pp. 623- 635, 2000.

....to the shape of the structuring element. The edge localization, Euclidean invariance, and causality properties of scale spaces generated by area morphology have been utilized in recent applications such as document page segmentation [2] image reconstruction [3] and image classification [1]. The major drawback of area morphology is the immense computational cost associated with the connected component analysis. In order to use area morphology in time critical applications, we have developed algorithms that can reduce the processing time for a typical area morphology operation by a ....

S.T. Acton and D.P. Mukherjee, "Scale space classification using area morphology," 1EEE Transactions on Image Processing, vol. 9, pp. 623-635, 2000.

....EDGES FROM AREA MORPHOLOGY oct T. Acton; and Dipti Prasad Mukherjee School of Electrical and Computer Engineering Oklahoma State University Stillwater, Oklahoma 74078 USA sacton okstate.edu Electronics and Com,nunications Sciences Unit Indian Statistical Institute India 700035 dipfi isical.ac.in ABSTRACT This paper introduces an edge detection process bascd on area morphology. Area open close and area close open operators are used to generate scaled ....

....EDGES FROM AREA MORPHOLOGY oct T. Acton; and Dipti Prasad Mukherjee School of Electrical and Computer Engineering Oklahoma State University Stillwater, Oklahoma 74078 USA sacton okstate.edu Electronics and Com,nunications Sciences Unit Indian Statistical Institute India 700035 dipfi isical.ac.in ABSTRACT This paper introduces an edge detection process bascd on area morphology. Area open close and area close open operators are used to generate scaled hnage representations for feature extraction. The edges are defined by the boundaries of the sealed objects in ....

[Article contains additional citation context not shown here]

S.T. Acton and D.P. Mukherjee, "Scale space classification using area morphology," accepted for publication, IEEE Trasactions on Image Processing.

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