Efficient computation of local geometric moments
Tóm tắt
Local moments have attracted attention as local features in applications such as edge detection and texture segmentation. The main reason for this is that they are inherently integral-based features, so that their use reduces the effect of uncorrelated noise. The computation of local moments, when viewed as a neighborhood operation, can be interpreted as a convolution of the image with a set of masks. Nevertheless, moments computed inside overlapping windows are not independent and convolution does not take this fact into account. By introducing a matrix formulation and the concept of accumulation moments, this paper presents an algorithm which is computationally much more efficient than convolving and yet as simple.
Từ khóa
#Image edge detection #Optical noise #Optical computing #Image segmentation #Noise reduction #Convolution #Image texture analysis #Polynomials #Geometrical optics #Nonlinear opticsTài liệu tham khảo
yaroslavsky, 1996, local adaptive image restoration and enhancement with the use of dft and dct in a running window, Proc SPIE, 2825, 1
10.1007/3-540-60268-2_359
10.1016/0167-8655(94)00089-L
10.1016/0031-3203(91)90047-9
10.1109/78.485943
10.1109/ROBOT.1993.291861
10.1016/0031-3203(91)90092-J
10.1109/83.370680
10.1016/0031-3203(92)90135-6
10.1109/34.41367
10.1109/70.134273
10.1109/TPAMI.1981.4767047
martinez, 1998, Accumulation moments Theory and applications
10.1109/34.3913
10.1109/ICIP.1994.413677
10.1002/(SICI)1099-1115(199603)10:2/3<283::AID-ACS351>3.3.CO;2-2
10.1016/0031-3203(94)00116-4
golub, 1996, Matrix Computations
10.1109/83.585230
10.1016/0167-8655(94)90069-8
10.1117/12.139804
10.1109/TASSP.1986.1164853
bronshtein, 1982, Handbook of Mathematics for Engineers, 158
10.1109/5.503295
10.1109/TIT.1962.1057692
10.1109/ROBOT.1996.506509
10.1142/9789812816092
martinez, 1998, fast actualization of moments in sliding window applications, Proc 9th Eur Signal and Image Processing Conf (EUSIPCO-98), 1, 145
reeves, 1982, a parallel mesh moment computer, Proc Int l Conf Pattern Recognition, 465
10.1109/34.85675
10.1109/ICASSP.1996.543671
10.1145/800260.809027