2007/8 Vol. J90 D No. 8 Stauffer [7] 2 2 I 1 I 2 2 (I 1(x),I 2(x)) 2 [13] I 2 = CI 1 (C >0) (I 1,I 2) (I 1,I 2) Field Monitoring Server

Size: px

Start display at page:

Download "2007/8 Vol. J90 D No. 8 Stauffer [7] 2 2 I 1 I 2 2 (I 1(x),I 2(x)) 2 [13] I 2 = CI 1 (C >0) (I 1,I 2) (I 1,I 2) Field Monitoring Server"

なごみはなだて
5 years ago
Views:

1 a) Change Detection Using Joint Intensity Histogram Yasuyo KITA a) 2 (0 255) (I 1 (x),i 2 (x)) I 2 = CI 1 (C>0) (I 1,I 2 ) (I 1,I 2 ) 2 1. [1] 2 [2] [3] [5] [6] [8] Intelligent Systems Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), AIST Tsukuba Central 2, Umezono, Tsukuba-shi, Japan a) y.kita@aist.go.jp [9] [11] 2 [12] 2 [6] [9] 1 D Vol. J90 D No. 8 pp c

2 2007/8 Vol. J90 D No. 8 Stauffer [7] 2 2 I 1 I 2 2 (I 1(x),I 2(x)) 2 [13] I 2 = CI 1 (C >0) (I 1,I 2) (I 1,I 2) Field Monitoring Server FMS [14], [15] (e) (f) (g) (h) 1 Fig. 1 Relation between subtraction processes and combinatorial intensity levels I 1(x) ai 2(x) a =( n I1(x)/n)/( n I2(x)/n) i=1 i=1 I k x n k (0 255)

3 30 2 1(e) 1 1(f) I 1I 2 2 I 1(x) (ai 2(x) 30) (ai 2(x) + 30) I 1I 2 1(g) (I 1,I 2) (I 1,I 2) 1(h) 1(e) 2 I 2 = I 1 C I 2 = CI 1 C C m I 1 I 2 N(C m,σ) I 1 I 2 = C mi 1 N(a, b) a b 2 1 I 2/I 1 I 2/I 1 =0.96 I 2 =0.96I 1 (I 1,I 2) 2 (I 1,I 2) 2 2 Fig. 2 Example of changed area based on combinatorial intensity levels I 1 I 2 N(C m,σ) I 1 1 I 2 =0.6I 1 C 3 I 2 = CI 1 (C >0)

0 4 Fig. 4 Changed Pixels obtained based inference of background clusters. 3.

4 2007/8 Vol. J90 D No. 8 3 Fig. 3 Diagram of distribution of various changes I II III (I 1,I 2) i 4 σ =3.0 1 I 1I 2 3 I 2 = CI 1 I 1I 2 (r, θ) r θ r = Fig. 4 Changed Pixels obtained based inference of background clusters. 3. r = I 1I 2 4 I 1I 2 ii L1 D (0, 0) (255, 255) iii (I 1,I 2) Sig(I 1,I 2) Sig(I 1,I 2)=0 if(i 1,I 2) is included in selected clusters. Sig(I 1,I 2)=1 else 1960

(I 1(x),I 2(x)) σ L1 D1 σ L1 D1 I 2 = C mi 1 4 Sig(I 1,I 2)=0 4 2.

5 (I 1(x),I 2(x)) σ L1 D1 σ L1 D1 I 2 = C mi 1 4 Sig(I 1,I 2)= N1 N1 N1 = 300 x y S e = min(e x,e y) n i=1 E x = (e1,x(x) µe 1,x)(e 2,x(x) µ e2,x ) n(max(σ e1,x,σ e2,x )) 2 n i=1 E y = (e1,y(x) µe 1,y )(e 2,y(x) µ e2,y ) n(max(σ e1,y,σ e2,y )) 2 e i,k µ ei,k σ ei,k i k S e 0.0 S e Fig. 5 Discrimination using gradient correlation. 0.1 < S e < S e S e T 1 T FMS σ L1 D (pixels) 30 (deg)

6 2007/8 Vol. J90 D No. 8 6 Fig. 6 Results examples under various conditions I 2/I <I 2/I 1 < 1.4 SIG(I 1,I 2)

7 2 I 2/I Sig(I 1,I 2)=0 1.0 ±3σ 7(e) ±3σ N (e) (f) 7 Fig. 7 Example of failure. (e) (f) 8 1 Fig. 8 Experimental results

8 2007/8 Vol. J90 D No Fig. 9 Experimental results (e) Sig(I 1,I 2) T 1=0 9 9 Intel/Xeon 2.4GHz 100 ms ms Bromiley [16] (I 1,I 2) (I 1,I 2) / 3 3 σ L1 D1 2 N1 S e

[1] R.J. Radke, S. Andra, O. Al-Kofahi, and B. Roysam, Image change detection algorithms: A systematic survey, IEEE Trans. Image Process., vol.14, no.3, pp.294 307, 2005. [2] B. Zitová and J.

9 [1] R.J. Radke, S. Andra, O. Al-Kofahi, and B. Roysam, Image change detection algorithms: A systematic survey, IEEE Trans. Image Process., vol.14, no.3, pp , [2] B. Zitová and J. Flusser, Image registration methods: A survey, Image Vis. Comput., vol.21, no.11, pp , [3] K. Toyama, J. Krumm, B. Brumit, and B. Meyers, Wallflower: Principles and practice of background maintenance, Proc. International Conference on Computer Vision, pp , [4] D-II vol.j86-d-ii, no.6, pp , June [5] M. Pic, L. Berthouze, and T. KuritaB, Adaptive background estimation: Computing a pixel-wise learning rate from local confidence and global correlation values, IEICE Trans. Inf. & Syst., vol.e87-d, no.1, pp.50 57, Jan [6] D-II vol.j79-d-ii, no.4, pp , April [7] C. Stauffer and E. Grimson, Adaptive background mixture models for real-time tracking, Proc. of Compurt Vision and Pattern Recognition 99, pp , [8] watershed D-II vol.j84-d-ii, no.12, pp , Dec [9] D-II vol.j84-d-ii, no.10, pp , Oct [10] vol.44, no.sig 5 (CVIMn 6), pp.54 63, [11] D-II vol.j87-d-ii, no.5, pp , May [12] Radial Reach Filter RRF D- II vol.j86-d-ii, no.5, pp , May [13] F. Maes, A. Collignon, D. Vandermeulen, G. Marchal, and P. Suetens, Multimodality image registration by maximization of mutual information, IEEE Trans. Med. Imaging, vol.16, no.2, pp , [14] FieldServerEn/default.htm [15] T. Fukatsu and M. Hirafuji, Field monitoring using sensor-nodes with a Web server, J. Robotics and Mechatronics, vol.17, no.2, pp , [16] P.A. Bromiley, N.A. Thacker, and P. Courtney, Nonparametric image subtraction using grey level scattergrams, Image Vis. Comput., vol.20, pp , Oxford 2005 IEEE CS 1965

4. C i k = 2 k-means C 1 i, C 2 i 5. C i x i p [ f(θ i ; x) = (2π) p 2 Vi 1 2 exp (x µ ] i) t V 1 i (x µ i ) 2 BIC BIC = 2 log L( ˆθ i ; x i C i ) + q

4. C i k = 2 k-means C 1 i, C 2 i 5. C i x i p [ f(θ i ; x) = (2π) p 2 Vi 1 2 exp (x µ ] i) t V 1 i (x µ i ) 2 BIC BIC = 2 log L( ˆθ i ; x i C i ) + q x-means 1 2 2 x-means, x-means k-means Bayesian Information Criterion BIC Watershed x-means Moving Object Extraction Using the Number of Clusters Determined by X-means Clustering Naoki Kubo, 1 Kousuke