R. Wilson and A. H. Bhalerao. Kernel Designs for Efficient Multiresolution Edge Detection and Orientation Estimation. IEEE Trans. PAMI, March, 1992.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....For discussions of steerable filters in three or more dimensions, see [59, 58, 33, 89] Martin Friedmann rendered this image with the Thingworld program) 33 Chapter 3 Analyzing Orientation 3. 1 Analyzing the DominantOrientation Orientation analysis is an important task in early vision [54, 57, 60, 117, 112]. Knutsson and Granlund [57] devised an elegantmethodforcombining the outputs of quadrature pairs to extract a measure of orientation. We describe a related method which makes optimal use of the filters designed in Section 2.4. We measure the orientation strength along a particular direction, ....

R. Wilson and A. H. Bhalerao. Kernel designs for efficientmultiresolution edge detection and orientation estimation. IEEE Pat. Anal. Mach. Intell., 14(3):384-- 390, 1992.

.... Rosenfeld JolRos94 book (1994) the books edited by Rosenfeld (1984) Cantoni and Levialdi (1986) and the special issue of IEEE TPAMI edited by Tanimoto (1989) A selection of recent developments can also be found in the articles by Chehikian and Crowley (1991) Knudsen and Christensen (1991) and Wilson and Bhalerao (1992), for a selection of recent developments. An interesting approach is the introduction of oversampled pyramids , in which not every smoothing step is followed by a subsampling operation, and hence, a denser sampling along the scale direction can be obtained. It is worth noting that pyramid ....

R. Wilson and A. H. Bhalerao, "Kernel design for efficient multiresolution edge detection and orientation estimation", IEEE Trans. Pattern Analysis and Machine Intell., vol. 14, no. 3, pp. 384--390, 1992.

....we are interested in high frequency features, e.g. texture, lines and edges. High pass filtering will emphasize such features. For efficiency reasons we implemented the high pass filter in terms of a Gaussian pyramid [27] using a near optimal 4 Theta 4 kernel developed by Wilson and Bhalerao [28]. We performed the multiresolution motion estimation algorithm on frame pairs over the high pass filtered sequence. The result from this is a quadtree description containing the estimated local motion parameters. For display purposes, we magnified the motion vectors and plotted them on top of the ....

R. Wilson and A. H. Bhalerao, "Kernel Designs for Efficient Multiresolution Edge Detection and Orientation Estimation," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 14, Mar. 1992.

.... depend on the scale of the objects therein [85, 38] For instance, several researchers have developed stereo matching techniques which operate at a range of scales, allowing objects to be matched despite changes in size [94, 115] Others have developed multiresolution methods for edge detection [147, 114, 93], filtering [32, 110] and compression [21, 44] Such multiresolution approaches often also lead to substantial computational advantages, especially when the processing is organized in a coarse to fine progression. For instance, in [18] a multiresolution segmentation algorithm is proposed in which ....

R. Wilson and A. Bhalerao. "Kernel design for efficient multiresolution edge detection and orientation estimation". IEEE Transactions on Pattern Analysis and Machine Intelligence, 14:384--390, 1992.

....state of boundary graph Figure 2: The segmentation algorithm consists of two main stages: initialisation followed by iterative processing and cooperative interaction between and . gorithm is based on the simplest multiresolution representation: the lowpass pyramid, 37] 38] 39] [40]. From the pyramid, a set of nodes is selected as forming an estimate of the model quadtree: stage 1) Links to neighbours are then defined in 6 a modified form of a region adjacency graph [41] 42] Each link has a binary state: on or off, which determines the membership ....

....impact on both the number of iterations to reach a steady state and, because the algorithm is deterministic, the quality of the final result. A robust initial estimate is therefore an important component of the algorithm. The first step is a smoothing operation performed using a lowpass pyramid [40] which trades off noise reduction against spatial resolution. Given an image # , and a lowpass kernel 7 , the general form of the processing is 9 F F 0 (4) where ....

[Article contains additional citation context not shown here]

R. Wilson and A. H. Bhalerao. Kernel Designs for Efficient Multiresolution Edge Detection and Orientation Estimation. IEEE Trans. PAMI, March, 1992.

....representations. It should be noted that this use of discrete rather than continuous forms poses problems in terms of the decimation and interpolation operations which implement the change of scale, in that a certain amount of aliasing can be introduced, leading to a lack of circular symmetry [34]. The multiresolution approach takes into account the requirement for invariance over scale by representing the image at several different spatial resolutions. This necessarily introduces an element of redundancy, but this is a relatively small price to pay for the advantages it affords. In this ....

....had a way of determining these local symmetries. A simple and elegant way of doing this would seem to be the use of a set of two dimensional filters to detect shape features. The technique used to detect shape features consists of applying a pair of 3 X 3 filters developed by Wilson and Bhalerao [34]. Each of these filters determine the first partial derivative of the image, and are oriented in mutually orthogonal directions. These orientations are at 45 degrees to the x and y axes, we shall call these axes the u and v axes. This is illustrated in Figure 5. These derivatives can then be ....

R. Wilson and A. H. Bhalerao. Kernel Designs for Efficient Multiresolution Edge Detection and Orientation Estimation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 14:384--390, 1992.

No context found.

R. Wilson and A. H. Bhalerao. Kernel design for efficient multiresolution edge detection and orientation estimation. IEEE Trans. Pattern Analysis and Machine Intelligence, 14(3):384--390, 1992.

No context found.

Wilson, R., and Bhalerao, A. H. 1992, "Kernel Designs for Efficient Multiresolution Edge Detection and Orientation Estimation", IEEE Trans. Patt. Anal. & Mach. Int., 14, 384-390.

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