Hatamian, M. (1986) "A Real-Time Two-Dimensional Moment Generating Algorithm and Its Single Chip Implementation," IEEETransacNJSL onAc---#U8JSL SpeecU and SignalProcJTT8--- , Vol. 34, No. 3, pp. 546--553, June.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....polyhedra, their algorithm proceeds by projecting the triangles of the boundary to the barycenter, and evaluates the integrals for the tetrahedra generated this 2 way. This method seems to be suitable, however we suspect that long, thin tetrahedra would cause numerical instability. Hatamian [8] presented an algorithm for efficient computation of arbitrarily high order moments in 2D discrete images which may easily be modified to handle 3D volumetric data. In this latter case an inaccuracy would be introduced by sampled nature of the data we suppose that this effect dramatically ....

M. Hatamian. A real-time two-dimensional moment generating algorithm and its single chip implementation. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1986.

....its standard position. 1 Introduction The purpose of this article is to put together all pieces concerning 3D Zernike moments to be introduced in this report, their fast computation and connection to the complete set of moment invariants using spherical harmonics [4] we had derived in [3] In [5] a fast algorithm and its single chip implementation for computing the ordinary geometrical moments of a 2D image has been presented that achieved a saving of more than 5 orders of magnitude regarding the number of multiplications needed compared to the direct implementation if the task is to ....

M. Hatamian, " A Real-Time Two-Dimensional Moment Generating Algorithm and its Single Chip Implementation" IEEE Trans. on Acoustics, Speech and Signal Processing, vol. ASSP-34, pp.546-553, June 1986.

....of horizontal to vertical lines in the symbol. Another set of characteristics are based on moments about the centroid of the blob. Moments have been found to be very useful in this type of pattern recognition [2,7] and a hardware implementation is possible allowing substantial speed improvements [6]. The recognition is done by comparing the feature vector of each blob found to those of prototype copies of the symbols we wish to match. These are all calculated offline in advance from a training set of board images. The simplest recogniser might just find the closest prototype in the ....

Hatamian, M., "A Real-Time Two-Dimensional Moment Generating Algorithm and Its Single Chip Implementation", IEEE Trans. on Acoustics, Speech, and Signal Processing, Vol ASSP-34, No. 3, June 1986, pp. 546-553

....theorem. One of them is used to compute geometric moments from binary image; the other to compute moments of regions in grey level images. 2 State of the art In order to speed up the computation of moments, the following techniques have been used. 1. Image ltering or transform: Hatamian [15] computed the moments by a causal spatial lter. The ltering needs only additions of O(N 2 ) for 2D images and O(N) for 1D signals. Hatamian developed an algorithm for computing the moments of grey level images. Fu et al. 21] found the relation between the moments and the coeOEcients of the ....

....1 arrays of size N . During contour following, the array entries are updated. Comparing equation (15) with (18) we have u ij = X y v i (y)y j (19) which means that u ij is the j th moment of a 1D signal v i (y) A fast algorithm for computing moments of a 1D signal was proposed by Hatamian [15]. Letting v 0 i (y) be the result of Hatamian ltering of v i (y) we have v 0 i (y) y X k=N v i (k) 20) Applying the Hatamian s lter recursively, we obtain v j i (y) y X k=N v j Gamma1 i (k) 21) Then the 1D moments u ij are linear combinations of v j i (1) For j 3 we ....

[Article contains additional citation context not shown here]

M. Hatamian, A real-time two-dimensional moment generating algorithm and its single chip implementation, IEEE Trans. ASSP 34(3), 546553(1986).

....by T (y; z) and de ne v p (y; z) as v p (y; z) X T(y;z) x p 1 Deltax (11) Comparing Eq. 8) with Eq. 11) we have u pqr = X y X z v p (y; z)y q z r (12) which means that u pqr is the (q r) th order moment of a 2 D gray level image with intensity function v p (y; z) Hatamian [3] proposed a method for fast computation of the geometric moments of a 2 D gray level image. This method uses a digital lter which we call the Hatamian lter. To compute the moments of a 2 D image, we rst lter the image in one direction and then the other direction. Letting v 0 p (y; z) be the ....

Hatamian, M.: A real-time two-dimensional moment generating algorithm and its single chip implementation. IEEE Trans. ASSP 34 (1986) 546553

No context found.

Hatamian, M. (1986) "A Real-Time Two-Dimensional Moment Generating Algorithm and Its Single Chip Implementation," IEEETransacNJSL onAc---#U8JSL SpeecU and SignalProcJTT8--- , Vol. 34, No. 3, pp. 546--553, June.

No context found.

M. Hatamian, A real-time two dimensional moment generating algorithm and its single chip implementation, IEEE Trans. ASSP 34 (1986) 546--553.

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