Ȩ > Ç¥ÁØÈ °³¿ä > TTAÀÇ Ç¥ÁØÇöȲ
Ç¥ÁعøÈ£ | TTAK.KO-10.0587 | ±¸Ç¥ÁعøÈ£ | |
---|---|---|---|
Á¦°³Á¤ÀÏ | 2012-10-09 | ÃÑÆäÀÌÁö | 16 |
ÇѱÛÇ¥Áظí | ¿µ»ó SoCÀÇ Æ¯Â¡Á¡ ÀúÀå Çü½Ä | ||
¿µ¹®Ç¥Áظí | Format for Storaging Feature Points of Vision SoC | ||
Çѱ۳»¿ë¿ä¾à | º» Ç¥ÁØÀº ÃÔ¿µµÈ ¿µ»ó¿¡¼ ÃßÃâµÈ ¿µ»óÀÇ Æ¯Â¡Á¡À» ÀúÀåÇÏ´Â Çü½ÄÀ» Á¤ÀÇÇÑ´Ù. ¿µ»ó SoC¿¡¼ ±âÁ¸ÀÇ ¿µ»ó ÀúÀåÀ» À§ÇÑ ¿µ»ó Çì´õ ¹× Çȼ¿ µ¥ÀÌÅÍ¿Í´Â º°µµ·Î Ư¡Á¡ Á¤º¸ Çì´õ°¡ Ãß°¡µÇ¾î »ç¿ëµÈ´Ù. Ư¡Á¡ Á¤º¸°¡ ÀúÀåµÉ Ư¡Á¡ Á¤º¸ Çì´õÀÇ Çü½ÄÀ» Á¤ÀÇÇϸç, Ư¡Á¡ Á¤º¸ Çì´õ¿¡ Ư¡Á¡/Ư¡ º¤ÅÍ Á¤º¸ÀÇ À¯¹«, »ý¼º ¹æ¹ý, ÀúÀå ¹æ¹ý, Ư¡Á¡/Ư¡º¤ÅÍ µ¥ÀÌÅÍ µîÀÇ Á¤º¸¸¦ ÀúÀåÇÑ´Ù. Ư¡Á¡ Á¤º¸ Çì´õ´Â Ư¡Á¡/Ư¡ º¤ÅÍ Á¤º¸ À¯¹« Ç¥½ÃÀÚ, Ư¡Á¡/Ư¡ º¤ÅÍ Á¤º¸ µ¥ÀÌÅÍ·Î ±¸¼ºµÇ¸ç, Ư¡Á¡/Ư¡ º¤ÅÍ Á¤º¸ µ¥ÀÌÅÍ¿¡´Â Ư¡Á¡ »ý¼º ¹æ¹ý, Ư¡Á¡ ÀúÀå ¹æ¹ý, Ư¡Á¡ µ¥ÀÌÅÍ, Ư¡Á¡ ±¸ºÐÀÚ, Ư¡ º¤ÅÍ »ý¼º ¹æ¹ý, Ư¡ º¤ÅÍ Å©±â, Ư¡ º¤ÅÍ ÀúÀå ¹æ¹ý, Ư¡ º¤ÅÍ µ¥ÀÌÅÍ, Ư¡ º¤ÅÍ ±¸ºÐÀÚ µîÀÌ ÀúÀå µÈ´Ù. ¿µ»óÀÇ Æ¯Â¡Á¡À» Ãʱ⿡ Çѹø¸¸ ÃßÃâÇÏ¿© À̸¦ ¿øº» ¿µ»ó¿¡ Ãß°¡ÇÏ¿© »ç¿ëÇÔÀ¸·Î½á Ư¡Á¡ ÃßÃâÀ» À§ÇÑ ¿¬»êÀ» ÁÙÀÏ ¼ö ÀÖ´Ù. | ||
¿µ¹®³»¿ë¿ä¾à | This standard defines the recording format of feature points in images. In vision SoC, feature point information header is used additionally apart from the previous image header and pixel data. The absence of feature points and feature vectors, generating method, storaging method and the data of feature points and feature vectors are saved in the feature point information header. The feature point information header consists of an absence-indicator of feature points ¡¤ feature vectors information and the data of feature points and feature vectors. In the data of feature points and feature vectors, generating method, storaging method of feature points, feature point data, feature point separator, feature vector generating method, storaging method, feature vector size, vector data and feature vector separator are saved. By extracting feature points only once in initial stage to additionally use in original feature, operation for extracting feature points can be reduced. | ||
±¹Á¦Ç¥ÁØ | |||
°ü·ÃÆÄÀÏ | TTAK_KO-10_0587.pdf |