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

»çÀÌÆ®¸Ê

Loading..

Please wait....

Çмú´ëȸ ÇÁ·Î½Ãµù

Ȩ Ȩ > ¿¬±¸¹®Çå > Çмú´ëȸ ÇÁ·Î½Ãµù > Çѱ¹Á¤º¸Åë½ÅÇÐȸ Çмú´ëȸ > 2019³â Ãß°èÇмú´ëȸ

2019³â Ãß°èÇмú´ëȸ

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) Convolution Neural Network¸¦ ÀÌ¿ëÇÑ µµ·ÎÀÎ½Ä ¸ðµ¨
¿µ¹®Á¦¸ñ(English Title) A Model for Road Recognition Using Convolution Neural Network
ÀúÀÚ(Author) ¾ç¿À¼®   ±è¿µ¿í   ÀÌÈ«·Î   ÀÌâ¿ì   Oh-seok Yang   Young-uk Kim   Hong-ro Lee   Chang-Woo Lee  
¿ø¹®¼ö·Ïó(Citation) VOL 23 NO. 02 PP. 0626 ~ 0628 (2019. 10)
Çѱ۳»¿ë
(Korean Abstract)
ÃÖ±Ù ¸î ³â µ¿¾È ÄÄÇ»ÅÍ ¿£Áö´Ï¾î¸µ Àü°øÀÚ¿Í ºñÀü°øÀڵ鿡°Ô ¸¹ÀÌ ¾ð±ÞµÇ´Â Àü°ø ´Ü¾î°¡ ÀΰøÁö´ÉÀ̶ó°í »ý°¢µÈ´Ù. °¡Æ®³Ê¿¡¼­´Â Áö³­ ¸î ³â µ¿¾È Áö´ÉÇüÀ̳ª AI¿Í °°Àº ÀΰøÁö´É°ú °ü·ÃµÈ ±â¼úÀ» ÁÖ¿ä ±â¼ú·Î »Ì¾Æ¿Ô´Ù. ±×¿¡ ¹ß¸ÂÃç Å×½½¶ó¿Í °°Àº ¸î¸î ÀÚµ¿Â÷ ȸ»çµéÀº ÀڽŵéÀÇ ÀÚµ¿Â÷¿¡ ¿ÀÅäÆÄÀÏ·µ ±â´ÉÀ» žÀçÇÏ¿© »ç¿ëÀڵ鿡°Ô ¼±º¸¿´´Ù. ¶ÇÇÑ, ÇöÀç µå·Ð¿¡´Â ÀÚÀ²ÁÖÇà ±â´ÉÀÌ Å¾ÀçµÈ ±â¼úÀÌ Áö¼ÓÇؼ­ ¿¬±¸µÇ°í ÀÖ´Ù. ¸¸ÀÏ µå·Ð¿¡ ÀÚÀ²ÁÖÇà ±â´ÉÀÌ Å¾ÀçµÈ´Ù¸é ¸¹Àº ºÐ¾ßÀÇ »ç¶÷µé¿¡°Ô ÀÌÁ¡À» Á¦°øÇÒ °ÍÀ¸·Î ±â´ëµÈ´Ù. º» ¿¬±¸´Â ÀÚÀ²ÁÖÇàÀ» µå·Ð¿¡ Àû¿ë °¡´ÉÇÑ µµ·ÎÀνÄÀ» À§ÇØ µö·¯´× ÇнÀ ¸ðµ¨·Î CNN(Convolution Neural Network)À» ÀÌ¿ëÇÏ¿© °³¹ßÇÑ´Ù. ÀÌ ÇнÀ ¸ðµ¨¿¡¼­ ¾ÕÀÇ 6°èÃþÀº Convolution ·¹À̾î·Î »ç¿ëÇÏ°í, µÚÀÇ 2°èÃþÀº Fully Connected ·¹À̾î·Î °³¹ßÇÏ°íÀÚ ÇÑ´Ù. º» ¿¬±¸´Â ÀÌ CNN ±â¹ýÀ» ÀÌ¿ëÇÏ¿© 935¡¿515 Å©±âÀÇ À̹ÌÁö¿¡ ´ëÇÑ µµ·ÎÀÎ½Ä Á¤È®µµ¸¦ °³¼±ÇÏ°íÀÚ ÇÑ´Ù.
¿µ¹®³»¿ë
(English Abstract)
In recent years, AI is thought to be the major word that is often mentioned by computer engineering majors and non-technicals. For the past few years, Gartner has tapped artificial intelligence-related technologies such as intelligence and AI as its main technology. In line with that, several automakers, such as Tesla, have introduced their cars to users with autopilot features. In addition, drone technology with self-driving capability is continuously being researched. If drones are equipped with self-driving functions, they are expected to provide benefits to people in many fields. This research is developed using CNN (Convolution Natural Network) as a deep learning model for road recognition that can be applied to drones. In this learning model, the first 6 layers will be used as the Convolution layer, while the 2 layer behind will be developed as the Fully Connected layer. This study uses this CNN technique to improve road recognition accuracy for images of 935¡¿515 size.
Å°¿öµå(Keyword) Deep Learning   CNN   Road Recognition   Image Processing  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå