HSI 500 SOM %89. RGB
HSI HSI Fig. (): The recognition block diagram HSI- RGB HSI RGB HSI. RGB HSI
HSI Fig. (): The result of applying HSI filter and extracting red colour Fig. (): The corner model 90 Fig. (4): The 90 degree corner mask X I(x, y) A if x > 0 and mx < y < mx = 0 Otherwise θ H = 60 θ = cos [ if if B <= G B > G [( R G ) + ( R B )] ( R G ) + ( R B )( G B )] [ min( R,G,B) ] ( R + G + B) / S = I = ( R + G + B) B G R RGB HSI H( x,y) ( ) < ( x,y) < Hmin < < H ifsmin < S x,y S b( x,y) = Imin < I Imax 0 otherwise H max H min max max b(x, y) I(x, y) S(x, y) H(x, y) H S I min min min = 0. = 0,, H S max max = 0.75 = 0. = 0, I max = 0.4 Detector Circular EdgesDCE Detector Radial EdgesDRE sobel
c c c 45 c 5 5 5 : c 4 p p c a p p c p, p a ± (a / ) p 4 p p 4 p c 5 c c 4 c c c c 5 p 5 n(x, y) m F(x,y)=I(x,y)+n(x,y) g(x,y) O(x, y) = F(x, y) g(x, y) g(x, y) g(x, y) mππ zw zw g(x, y) = a sin [ (e + e )] w nπx n πy g(x, y) = a sin sin w w w. a m,z,n m =,n, a w x n = n = z a a 9 9 90 90 554590 5
Fig. (8): The different stages of recognition circular signs 00 00 Bicubic 4 4 (weighted average ( Fig. (5): The general configuration of applying five masks to the image RWB RWB %9 c c Fig. (6): (a) The result of applying mask c (b) The result of applying mask c 545 55 Fig. (7): (a) The linear structure under 45 and 5 degrees (b) The linear structure under 5 and 5 degrees (c) The prependicular linear structure
08060400 Matlab Pentium-4 GH 55 055 cross correlation Fig. (0): Samples of affecting normalized cross correlation coefficient between velocity table numbers and data base numbers HSI RWB Fig. (9): (a) The extracted image in HSI colour space (b) Applying RWB filter (c) Noise reduction f. (N x,n y ) t (M x,m y ) 0 u M + M 0 v N + N [f(x, y) fu,v][t(x u,y v) t] x,y r(u, v) = [ [f(x, y)f u,v] [t(x u,y v) t] } x,y x,y x x y y t t f f (x, y) f u,v
[] Fig. (): (a) The image with very objects (b) The sign location by Ref [] method (c) The sign location by proposed method Table (): The rate of sign location error, for different noises and different distances for the proposed method and Ref [] method [] %0 %0 %0 %7 %9 %4 %5 % %7 %5 %0 % 0 0 5 0 % %7 %4 %0 %8 %9 % % 5 %0 %0 []
75 00505 0 0005 5 Fig. (): (a) The images of different sizes for different distances (from right to left) 5, 0, 0 and 0 meters (b) The images with different noise percents (from right to left) 5, 0, 0 and 0 percent at 0 meters distance (c) The rotation of image for 5 degrees in clockwise and anti-clockwise direction Table (): Time spent for proposed algorithm.4 0005 0 0 05 700 8 00505 5 5 60 %89 %9 %86
40 Km 60 Km 00 05 Fig. (): (a) The 5 meters (b) 0 meters (c) Long distance (d) 0 meters and 0 percent noise [] R. Malik, J. Khurshid, S.N. Ahmad, "Road sign detection and recognition using coloursegmentation, shape analysis and template matching", IEEE/ICMLC, Vol.6, pp.556 560, Hong Kong, Aug. 007. [] J. Miura, T. Kanda, Y. Shirai, "An active vision system for on-line traffic sign recognition", IEICE Trans., pp.784 79, 00. [] S.M. Prieto, R. Alastair Allen, "Using self-organising maps in the detection and recognition of road Signs", Sch. of Engi. Ima. and Visi. Comp., Vol.7, pp.67 68, 009. [4] M. Lalonde, Y. Ling, "Road sign recognition, survey of the state of the art", CRIM/IIT, 995. [5] J. Torresen, J.W. Bakke, L. Sekania, "Efficient recognition of speed limit signs", IEEE/ITS, Washington DC, Oct. 004. [6] P. Suau, "Robust artificial landmark recognition using polar histograms", In Lec. Not. in Comp. Sci., Vol.80, pp.455 46, 005. [7] Y. Ishizuka, E.A. Puente, "Segmentation of road signsymbolsusingopponent-colorfilters", ITSWC, Nagoya, pp.- 5, Oct. 004. [8] M.A. Garcia-Garrido, M.A. Sotelo, E. Martm-Gorostiza, "Fast traffic sign detection and recognition under changing lighting conditions", IEEE/ITSC, pp.8 86, Sep. 006. [9] H. Liu, D. Liu, J. Xin, "Real-time recognition of road traffic sign in motion image based on genetic algorithm", IEEE/ICMLC, Vol., pp.8-86,00. [0] H. Zhang, D. Luo, "A new method for traffic signs classification using probabilistic neuralnetworks", In: Wang, J. Yi, Z. Zurada, J.M. Lu, B.L. Yin, H. (eds.), ISNN 006. LNCS, Vol.97, pp. 9, 006.
[] Y.Y. Nguwi, A.Z. Kouzani, "Automatic road sign recognition using neural networks", IEEE/IJCNN, pp.955 96, 006. [] C.Y. Fang, S.W. Chen, C.S. Fuh, "Road sign detection and tracking", IEEE Trans. on Vehi. Tech., Vol.5, pp.9 4, 00. [] D.L. Escalera, L. Moreno, "Road traffic sign detection and classification", IEEE Trans. on Indu. Elec., Vol.44, No.6, pp.46-6, 997. [4] H. Sandoval, T. Hattori, S. Kitagawa, Y. Chigusa, "Angle dependent edge detection for traffic sign recognition", IEEE/IVS, pp.08-, 000. [5] K. Rangarajan, M. Shah, D. Van Brackle, "Optimal corner detector", Comp. Visi., Grap. and Ima. Proc., Vol.48, No., pp.0-45, Nov. 989.