NARL In the case of promotingFig turning over reverse side technique for recognizing whether promoted or not In the game promoting c recognition method for promoting or not general ID: 831915
Download Pdf The PPT/PDF document "recognize the motions than "simple-move"" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
recognize the motions than "simple-move"
recognize the motions than "simple-move" ("NARL"') In the case of "promoting"(Fig. turning over reverse side, technique for recognizing whether promoted or not. (In the game, promoting (c) recognition method for promoting or not general-purpose character used. If this method was used, easily. However, resolution of the obtained spread (Fig. computation cost KIFU real time would propose a simple recognizing "promoting" the difference the number black pixels the front reverse side on pieces In this method, "promoting" is the difference the number the character between before If the number of black after moving decreases the piece having turned to the reverse side, and recognized the piece is recognized that it being on front method, because computation cost a real-time pos
sible. Since is recognized count the bla
sible. Since is recognized count the black pixels a threshold the method based on the and least criteria''] for each square. Though, since each square, the number of black the same piece varies "promoting" becomes This paper describes a method that threshold for binarization each square from the existing known that the black pixels lund pieces become the Thresholds for Binarization Due to the intensity of shadow and on; the piece varies from square on a board. Our technique utilizes fact that the kind current piece each square all the have been traced from the beginning. Threshold Adjusting a square the gray level histogram for each square where a exists. Then, as the the value that this histogram value which frorrl being used for black pixels on the piece computes the ave
rage number pixels for kind of the targe
rage number pixels for kind of the target black pixels of the kind of the piece. it adjusts of the threshold values for the method B, in initial state thresholds are decided the threshold adjusting method Section 4.1 and use fixed value in subsequent times. is proposed entire method. that the proposed method is the variation pixels the said that proposal method correct recognizing automatic recording system have proposed a simple the difference the number the character between on the front side the reverse each piece. also proposed a method and the method grasping the environmental change calculation cost. suitable thresholds each square combining them. experimental results. [I] H. Kakiuchi, T. Fukuyama, A. Doumae, J. Yi Okazaki, "Automatic Image Processing," PRMU200 1-14
9, Technical Report of IEICE, Vol.101,
9, Technical Report of IEICE, Vol.101, N. Otsu, "An Automatic Threshold Selection Method Based on Discriminant Electronics, Information Communication Engineers, Vol.J63-D, No.4, April, 1980. MVA2002 IAPR Workshop on Machine Vision Applications, Dec. 11 - Nara, Japan 3-4 Automatic Recording System of the Plays and Moves SHOGI Games (naru), This paper focuses on "promoting". a simple for recognizing "promoting" automatic determination binarization utilizing peculiar to for recognizing "promoting". exist research for KIFU'". However, no simple method "simple-move"(ugoku), "promoting"(naru). "dropping-move"(utsu) and "capture-moveW(toru). In Recording the picture and black pixels of each square placed, the system knows each square board each time, the system