• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö > Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö ¼ÒÇÁÆ®¿þ¾î ¹× µ¥ÀÌÅÍ °øÇÐ

Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö ¼ÒÇÁÆ®¿þ¾î ¹× µ¥ÀÌÅÍ °øÇÐ

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) ¹«°ÔÁ߽ɰú Á¤Á¡ °£ÀÇ °Å¸® Ư¼ºÀ» ÀÌ¿ëÇÑ »ï°¢Çü ¸Þ½¬ÀÇ Á¤·Ä
¿µ¹®Á¦¸ñ(English Title) The Alignment of Triangular Meshes Based on the Distance Feature Between the Centroid and Vertices
ÀúÀÚ(Author) ±¸¹ÎÁ¤   Á¤»óÈÆ   ±è±¸Áø   Minjeong Koo   Sanghun Jeong   Ku-Jin Kim  
¿ø¹®¼ö·Ïó(Citation) VOL 11 NO. 12 PP. 0525 ~ 0530 (2022. 12)
Çѱ۳»¿ë
(Korean Abstract)
µÎ °³ÀÇ Á¡±º(point cloud)À» Á¤·Ä(alignment)Çϱâ À§ÇØ ÇöÀç±îÁö ICP(iterative closest point) ¾Ë°í¸®ÁòÀÌ ³Î¸® »ç¿ëµÇ°í ÀÖÁö¸¸, ICP´Â µÎ Á¡±ºÀÇ Ãʱ⠹æÇâÀÌ Å©°Ô ´Ù¸¦ °æ¿ì Á¤·Ä¿¡ ½ÇÆÐÇÏ´Â °æ¿ì°¡ ¸¹´Ù. º» ³í¹®¿¡¼­´Â µÎ °³ÀÇ »ï°¢Çü ¸Þ½¬ A, B°¡ ¼­·Î Å©°Ô ´Ù¸¥ Ãʱ⠹æÇâÀ» °¡Áú ¶§, À̵éÀ» Á¤·ÄÇÏ´Â ¾Ë°í¸®ÁòÀ» Á¦¾ÈÇÑ´Ù. ¸Þ½¬ A, B¿¡ ´ëÇØ °¢°¢ °¡ÁßÄ¡ ¹«°ÔÁß½É(weighted centroid)À» ±¸ÇÑ µÚ, ¹«°ÔÁß½ÉÀ¸·ÎºÎÅÍ Á¤Á¡±îÁöÀÇ °Å¸®¸¦ ÀÌ¿ëÇÏ¿© ¸Þ½¬ °£¿¡ ¼­·Î ´ëÀÀµÉ °¡´É¼ºÀÌ ÀÖ´Â Á¤Á¡µéÀ» Ư¡Á¡À¸·Î ¼³Á¤ÇÑ´Ù. ¼³Á¤µÈ Ư¡Á¡µéÀÌ ´ëÀÀµÉ ¼ö ÀÖµµ·Ï ¸Þ½¬ B¸¦ ȸÀüÇÑ µÚ, A¿Í BÀÇ Á¤Á¡µé¿¡ ´ëÇØ RMSD(root mean square deviation)¸¦ ÃøÁ¤ÇÑ´Ù. RMSD°¡ ±âÁØÄ¡º¸´Ù ÀÛÀº °ªÀ» °¡Áú ¶§±îÁö Ư¡Á¡À» º¯°æÇÏ¸ç °°Àº °úÁ¤À» µÇÇ®ÀÌÇÏ¿© Á¤·ÄµÈ °á°ú¸¦ ¾ò´Â´Ù. ½ÇÇèÀ» ÅëÇØ ICP ¹× Go-ICP ¾Ë°í¸®ÁòÀ¸·Î Á¤·ÄÀÌ ½ÇÆÐÇÒ °æ¿ì¿¡µµ Á¦¾ÈµÈ ¾Ë°í¸®ÁòÀ¸·Î Á¤·ÄÀÌ °¡´ÉÇÔÀ» º¸ÀδÙ.
¿µ¹®³»¿ë
(English Abstract)
Although the iterative closest point (ICP) algorithm has been widely used to align two point clouds, ICP tends to fail when the initial orientation of the two point clouds are significantly different. In this paper, when two triangular meshes A and B have significantly different initial orientations, we present an algorithm to align them. After obtaining weighted centroids for meshes A and B, respectively, vertices that are likely to correspond to each other between meshes are set as feature points using the distance from the centroid to the vertices. After rotating mesh B so that the feature points of A and B to be close each other, RMSD (root mean square deviation) is measured for the vertices of A and B. Aligned meshes are obtained by repeating the same process while changing the feature points until the RMSD is less than the reference value. Through experiments, we show that the proposed algorithm aligns the mesh even when the ICP and Go-ICP algorithms fail.
Å°¿öµå(Keyword) Á¤·Ä   ¸Þ½¬   °Å¸® Ư¼º   ¹«°ÔÁ߽ɠ  ICP   Go-ICP   Alignment   Mesh   Distance Feature   Centroid   ICP   Go-ICP  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå