µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)
Current Result Document : 14 / 14
ÇѱÛÁ¦¸ñ(Korean Title) |
´ë¿ë·® À̹ÌÁö µ¥ÀÌÅͺ£À̽º¸¦ À§ÇÑ ½ºÄÉÀϸµ-ºÒº¯ À±°û¼± À̹ÌÁö ¸ÅĪ ½Ã½ºÅÛÀÇ ¼³°è ¹× ±¸Çö |
¿µ¹®Á¦¸ñ(English Title) |
Design and Implementation of a Scaling-Invariant Boundary Image Matching System for Large Image Databases |
ÀúÀÚ(Author) |
±è¹ü¼ö
¹®¾ç¼¼
Bum-Soo Kim
Yang-Sae Moon
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 28 NO. 02 PP. 0055 ~ 0064 (2012. 08) |
Çѱ۳»¿ë (Korean Abstract) |
º» ³í¹®¿¡¼´Â ½Ã°è¿ ¸ÅĪ ±â¼úÀ» È°¿ëÇÑ ½ºÄÉÀϸµ-ºÒº¯ À±°û¼± À̹ÌÁö ¸ÅĪ ½Ã½ºÅÛÀ» ¼³°èÇÏ°í ±¸ÇöÇÑ´Ù. ½ºÄÉÀϸµ-ºÒº¯ÀÇ Áö¿øÀº À±°û¼± À̹ÌÁö ¸ÅĪ¿¡¼ º¸´Ù Á÷°üÀûÀÎ °á°ú¸¦ ¾ò±â À§ÇØ Áß¿äÇÑ ¿ä¼ÒÀÌ´Ù. ±âÁ¸ ¿¬±¸¿¡ µû¸£¸é, À±°û¼± À̹ÌÁö¸¦ ½Ã°è¿·Î ³ªÅ¸³¾ °æ¿ì, ½ºÄÉÀÏµÈ À¯»ç À̹ÌÁöµéÀ» ã´Âµ¥ °Å¸® °è»êÀÌ ¿ëÀÌÇØÁö°í, À妽º »ç¿ëÀÌ °¡´ÉÇÏ¿© ´ë¿ë·® µ¥ÀÌÅͺ£À̽º ´ë»óÀÇ ºü¸¥ °Ë»öÀÌ °¡´ÉÇÏ°Ô µÈ´Ù. º» ³í¹®¿¡¼´Â ±âÁ¸ ¿¬±¸¸¦ ¹ÙÅÁÀ¸·Î, »ç¿ëÀÚ ÆíÀǸ¦ À§ÇÑ GUI ȯ°æÀÇ Å¬¶óÀ̾ðÆ®-¼¹ö ½Ã½ºÅÛÀ¸·Î ½ºÄÉÀϸµ-ºÒº¯ À̹ÌÁö ¸ÅĪ ±â´ÉÀ» ¼³°è ¹× ±¸ÇöÇÑ´Ù. ½ÇÇè °á°ú, ±¸ÇöÇÑ ½Ã½ºÅÛÀº Á÷°üÀûÀÌ°í Á¤È®ÇÑ ¸ÅĪÀ» ¼öÇàÇÏ´Â °ÍÀ¸·Î ³ªÅ¸³µ´Ù. ¶ÇÇÑ, ±¸ÇöÇÑ À妽º ±â¹Ý ¸ÅĪÀº ¼øÂ÷ ¸ÅĪ¿¡ ºñÇØ ±× ¼º´ÉÀ» ¼ö ¹è¿¡¼ ¼ö½Ê ¹è±îÁö Çâ»ó½ÃÅ´À» ¼º´É ½ÇÇèÀ» ÅëÇØ È®ÀÎÇÏ¿´´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
In this paper we design and implement a scaling‐invariant boundary image matching system by exploiting time‐series matching techniques. Supporting scaling invariance is an important factor to achieve the more intuitive matching results in boundary image matching. According to the previous work, converting boundary images to time‐series makes it easier to compute distances for searching the scaled similar images, and this makes it feasible to perform fast search using indexes even on a large image database. Based on the previous work, in this paper we design and implement a client‐server system that supports the scaling‐invariant image matching function with very convenient GUI interfaces. Experimental results show that our system provides much intuitive and accurate matching result. In addition, compared with the simple sequential scan, our index‐based matching method improves performance by one or two orders of magnitude.
|
Å°¿öµå(Keyword) |
½Ã°è¿ µ¥ÀÌÅͺ£À̽º
µ¥ÀÌÅÍ ¸¶ÀÌ´×
À̹ÌÁö ¸ÅĪ
½ºÄÉÀϸµ-ºÒº¯
À¯»ç ½ÃÄö½º ¸ÅĪ
Time‐Series Matching
Data Mining
Boundary Image Matching
Scaling Invariance
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|