Çѱ¹ÀÎÅͳÝÁ¤º¸ÇÐȸ ³í¹®Áö
Current Result Document : 1 / 1
ÇѱÛÁ¦¸ñ(Korean Title) |
Á¦½ºÃ³ ÀνÄÀ» À§ÇÑ ¼Õ°¡¶ô °³¼ö ÀÎ½Ä ¹æ¹ý |
¿µ¹®Á¦¸ñ(English Title) |
A Finger Counting Method for Gesture Recognition |
ÀúÀÚ(Author) |
À̵µ¿±
½Åµ¿±Ô
½Åµ¿ÀÏ
DoYeob Lee
DongKyoo Shin
DongIl Shin
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 17 NO. 02 PP. 0029 ~ 0037 (2016. 04) |
Çѱ۳»¿ë (Korean Abstract) |
Àΰ£Àº ÀÇ»ç¼ÒÅëÀ» ÅëÇؼ »óÈ£°ü°è¸¦ À¯Áö½ÃÅ°°í ¹ßÀü½ÃÄѳª°£´Ù. ÀÇ»ç¼ÒÅëÀº Å©°Ô ¾ð¾îÀû ÀÇ»ç¼ÒÅë°ú ºñ¾ð¾îÀû ÀÇ»ç¼ÒÅëÀ¸·Î ³ª´¶´Ù. ¾ð¾îÀû ÀÇ»ç¼ÒÅëÀº ¸» ¶Ç´Â ±ÛÀ» »ç¿ëÇÏ´Â °ÍÀÌ°í ºñ¾ð¾îÀû ÀÇ»ç¼ÒÅëÀº ¸öµ¿ÀÛÀ¸·Î Àǻ縦 Àü´ÞÇÏ´Â °ÍÀÌ´Ù. ¿ì¸®´Â ÀÏ»ó»ýÈ°¿¡¼ ´ëȸ¦ ÇÒ ¶§ ¸»°ú ´õºÒ¾î Á¦½ºÃ³¸¦ ÇÔ²² »ç¿ëÇÑ´Ù. Á¦½ºÃ³´Â ºñ¾ð¾îÀû ÀÇ»ç¼ÒÅë¿¡ ¼ÓÇϸç, ´Ù¾çÇÑ ÇüÅÂ¿Í ¿òÁ÷ÀÓÀ¸·Î Àǻ縦 Àü´ÞÇÒ ¼ö ÀÖ´Ù. ÀÌ·¯ÇÑ ÀÌÀ¯·Î Á¦½ºÃ³´Â HCI ºÐ¾ß¿Í HRI ºÐ¾ß¿¡¼ NUI/NUX¸¦ ±¸ÇöÇϱâ À§ÇÑ ¼ö´ÜÀ¸·Î °¢±¤¹Þ°í ÀÖ´Ù. º» ³í¹®¿¡¼´Â Å°³ØÆ®¿Í ¼ÕÀÇ ±âÇÏÇÐÀûÀΠƯ¡À» »ç¿ëÇÏ¿© ¼Õ ¿µ¿ª °ËÃâ°ú ¼Õ°¡¶ô °³¼ö¸¦ ÀνÄÇÏ´Â ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. Å°³ØÆ®°¡ Á¦°øÇÏ´Â ±íÀÌ ¿µ»óÀ» ÀÌ¿ëÇÏ¿© ¿µ»ó¿¡¼ ¼Õ ¿µ¿ªÀ» °ËÃâÇÏ°í ¼ÕÀÇ À±°û¼±°ú ÁßÁ¡ÀÇ °Å¸®¸¦ ºñ±³ÇÏ¿© ¼Õ°¡¶ô °³¼ö¸¦ ÆľÇÇÑ´Ù. º» ³í¹®¿¡¼ Á¦¾ÈÇÑ ¹æ¹ý¿¡ µû¸¥ ¼Õ°¡¶ô °³¼ö ÀνķüÀº Æò±Õ 98.5%ÀÌ°í ¼öÇà½Ã°£Àº 0.065msÀÌ´Ù. ÀÌ ¹æ¹ýÀº ±âÁ¸ÀÇ ¿¬±¸¿Í ºñ±³ ÇßÀ» ¶§, ÀÎ½Ä ¼Óµµ°¡ ºü¸£¸ç, º¹Àâµµ°¡ O(n),À¸·Î½á ¼º´É ¶ÇÇÑ ¿ì¼öÇÏ´Ù. ÇâÈÄ À̸¦ ÅëÇØ Á¦½ºÃ³ÀÇ ÀÎ½Ä °¡´ÉÇÑ ¹üÀ§¸¦ Áõ°¡½ÃÄÑ º¸´Ù ÄÄÇ»ÅÍ¿Í Àΰ£ÀÇ »óÈ£ÀÛ¿ëÀÌ ¼ö¿ùÇØÁö´Âµ¥ µµ¿òÀÌ µÉ °ÍÀÌ´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
Humans develop and maintain relationship through communication. Communication is largely divided into verbal communication and non-verbal communication. Verbal communication involves the use of a language or characters, while non-verbal communication utilizes body language. We use gestures with language together in conversations of everyday life. Gestures belong to non-verbal communication, and can be offered using a variety of shapes and movements to deliver an opinion. For this reason, gestures are spotlighted as a means of implementing an NUI/NUX in the fields of HCI and HRI. In this paper, using Kinect and the geometric features of the hand, we propose a method for recognizing the number of fingers and detecting the hand area. A Kinect depth image can be used to detect the hand region, with the finger number identified by comparing the distance of outline and the central point of a hand. Average recognition rate for recognizing the number of fingers is 98.5%, from the proposed method, The proposed method would help enhancing the functionality of the human computer interaction by increasing the expression range of gestures.
|
Å°¿öµå(Keyword) |
Å°³ØÆ®
¼Õ Á¦½ºÃ³
¼Õ Á¦½ºÃ³ ÀνÄ
±íÀÌ ¿µ»ó
Kinect
Hand Gesture
Hand Gesture Recognition
Depth Image
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|