ÇѱÛÁ¦¸ñ(Korean Title) |
½Ç½Ã°£ °´Ã¼ ÃßÀûÀ» À§ÇÑ Condensation ¾Ë°í¸®Áò°ú Mean-shift ¾Ë°í¸®ÁòÀÇ °áÇÕ |
¿µ¹®Á¦¸ñ(English Title) |
Integration of Condensation and Mean-shift algorithms for real-time object tracking |
ÀúÀÚ(Author) |
Á¶»óÇö
°ÇàºÀ
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 12-B NO. 03 PP. 0273 ~ 0282 (2005. 06) |
Çѱ۳»¿ë (Korean Abstract) |
½Ç½Ã°£ °´Ã¼ ÃßÀû(Real-time object tracking)Àº ºñµð¿À °¨½Ã ½Ã½ºÅÛ, ºñÀü ±â¹Ý ³×ºñ°ÔÀÌÅÍ¿Í °°Àº ºñÀü ÀÀ¿ë »ê¾÷ÀÌ ¹ß´ÞÇÏ¸é¼ ±× Á߿伺ÀÌ ´õÇØÁö°í ÀÖ´Â ºÐ¾ßÀÌ´Ù. °´Ã¼ ÃßÀûÀ» À§ÇØ ¸¹ÀÌ ÀÌ¿ëµÇ°í ÀÖ´Â ¾Ë°í¸®ÁòÀ¸·Î Mean-shift¿Í Condensation ¾Ë°í¸®ÁòÀÌ ÀÖ´Ù. Mean-shift ¾Ë°í¸®ÁòÀ» ±â¹ÝÀ¸·Î ÇÑ °´Ã¼ ÃßÀû ¾Ë°í¸®ÁòÀº ±¸ÇöÀÌ °£´ÜÇÏ°í, ÀûÀº °è»ê º¹Àâµµ¸¦ °®´Â ÀåÁ¡ÀÌ ÀÖ´Ù. µû¶ó¼ ½Ç½Ã°£ °´Ã¼ ÃßÀû ½Ã½ºÅÛ¿¡ ÀûÇÕÇÏ´Ù°í ÇÒ ¼ö ÀÖÁö¸¸, Áö¿ª ¸ðµå(Local mode)·Î ¼ö·ÅÇϴ Ư¼ºÀ¸·Î ÀÎÇØ º¹ÀâÇÑ È¯°æ(Cluttered environment)¿¡¼´Â ÁÁÀº ¼º´ÉÀ» ³ªÅ¸³»Áö ¸øÇÏ´Â ´ÜÁ¡À» °¡Áö°í ÀÖ´Ù. ¹Ý¸é, ¿©·¯ °³ÀÇ È帵éÀ» ÀÌ¿ëÇØ °´Ã¼ÀÇ À§Ä¡¸¦ ÃßÁ¤ÇÏ´Â Condensation ÃßÀû ¾Ë°í¸®ÁòÀº º¹ÀâÇÑ È¯°æ¿¡¼ ƯÁ¤ °´Ã¼¸¦ ÃßÀûÇϴµ¥ ¸¹ÀÌ »ç¿ëµÈ´Ù. ÇÏÁö¸¸ Condensation ¾Ë°í¸®ÁòÀ» ±â¹ÝÀ¸·Î ÇÑ ÃßÀû ¾Ë°í¸®ÁòÀº Á¤È®ÇÑ ÃßÀûÀ» Çϱâ À§Çؼ º¹Àâµµ°¡ ³ôÀº °´Ã¼ ¸ðµ¨°ú ¸¹Àº ¼öÀÇ È帰¡ ¿ä±¸µÈ´Ù. µû¶ó¼ ³ôÀº º¹Àâµµ¸¦ °®°Ô µÇ°í, ÀÌ°ÍÀ¸·Î ÀÎÇØ º¹ÀâÇÑ È¯°æ¿¡¼´Â ½Ç½Ã°£ ±¸ÇöÀÌ ¾î·Æ´Ù´Â ´ÜÁ¡À» °®°Ô µÈ´Ù. º» ³í¹®¿¡¼´Â, º¹ÀâÇÑ È¯°æ¿¡¼ ½Ç½Ã°£ °´Ã¼ ÃßÀû¿¡ ÀûÇÕÇϵµ·Ï Condensation ¾Ë°í¸®Áò°ú Mean-shift ¾Ë°í¸®ÁòÀ» °áÇÕÇؼ, ÀûÀº ¼öÀÇ È帵éÀ» ÀÌ¿ëÇÏ´Â ¸ðµ¨À» Á¦¾ÈÇÑ´Ù. ÀûÀº ¼öÀÇ È帵éÀ» ÀÌ¿ëÇÏ´õ¶óµµ, Mean-shift ¾Ë°í¸®ÁòÀ» ÀÌ¿ëÇØ º¸´Ù ³ôÀº À¯»çµµ¸¦ °¡Áö´Â È帵鸸À» ÀÌ¿ëÇÔÀ¸·Î½á, Condensation ¾Ë°í¸®ÁòÀ̳ª Mean-shift ¾Ë°í¸®Áò¸¸À» ÀÌ¿ëÇÒ ¶§º¸´Ù ´õ ³ªÀº ¼º´ÉÀ» ¾òÀ» ¼ö ÀÖ¾ú´Ù. |
¿µ¹®³»¿ë (English Abstract) |
Real-time Object tracking is an important field in developing vision applications such as surveillance systems and vision based navigation. mean-shift algorithm and Condensation algorithm are widely used in robust object tracking systems. Since the mean-shift algorithm is easy to implement and is effective in object tracking computation, it is widely used, especially in real-time tracking systems. One of the drawbacks is that it always converges to a local maximum which may not be a global maximum. Therefore, in a cluttered environment, the Mean-shift algorithm does not perform well. On the other hand, since it uses multiple hypotheses, the Condensation algorithm is useful in tracking in a cluttered background. Since it requires a complex object model and many hypotheses, it contains a high computational complexity. Therefore, it is not easy to apply a Condensation algorithm in real-time systems. In this paper, by combining the merits of the Condensation algorithm and the mean-shift algorithm, we propose a new model which is suitable for real-time tracking. Although it uses only a few hypotheses, the proposed method use a high-likelihood hypotheses using mean-shift algorithm. As a result, we can obtain a better result than either the result produced by the Condensation algorithm or the result produced by the mean-shift algorithm., |
Å°¿öµå(Keyword) |
Mean shift
Condensation
Particle Filter
°´Ã¼ ÃßÀû
Real time Tracking
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|