2019³â Ãß°èÇмú´ëȸ
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
µö·¯´× ±â¹Ý Â÷·® ¼Óµµ ¸ð´ÏÅ͸µÀ» ÅëÇÑ ½º¸¶Æ® Ⱦ´Üº¸µµ ½Ã½ºÅÛÀÇ ¼³°è ¹× ±¸Çö |
¿µ¹®Á¦¸ñ(English Title) |
Design and Implementation of A Smart Crosswalk System based on Monitoring Vehicle Speed using Deep Learning |
ÀúÀÚ(Author) |
Àå¼±Çý
Á¶ÈñÀº
Á¤Áø¿ì
Sun-Hye Jang
Hee-Eun Cho
Jin-Woo Jeong
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 23 NO. 02 PP. 0522 ~ 0525 (2019. 10) |
Çѱ۳»¿ë (Korean Abstract) |
¿ì¸®³ª¶óÀÇ °³Àκ° ÀÚµ¿Â÷ º¸±Þ·üÀÌ Áö¼ÓÀûÀ¸·Î Áõ°¡ÇÏ¸é¼ ±³Åë»ç°í ¹ß»ý·ü ¶ÇÇÑ Áö¼ÓÀûÀ¸·Î Áõ°¡ÇÏ´Â Ãß¼¼¸¦ º¸ÀÌ°í ÀÖ´Ù. ƯÈ÷, º¸ÇàÀÚ »ç°í Áß È¾´Üº¸µµ¿¡¼ÀÇ ¾Æµ¿°ú ³ëÀÎÀÇ »ç°í ºñÀ²ÀÌ ³ôÀº °ÍÀ¸·Î ³ªÅ¸³ª¸é¼ Ⱦ´Üº¸µµ ±³Åë»ç°í¿¡ ´ëÇÑ ¿¹¹æ ³ë·ÂÀÌ ¸¹ÀÌ ÀÌ·ç¾îÁö°í ÀÖ´Ù. º» ³í¹®¿¡¼´Â Ⱦ´Üº¸µµ¿¡ Á¢±ÙÇÏ´Â Â÷·®µéÀÇ ¼Óµµ ¸ð´ÏÅ͸µÀ» ÅëÇÑ ÀûÀÀÇü ¾Ë¸²À» Á¦°øÇÔÀ¸·Î½á º¸ÇàÀÚ ±³Åë»ç°í¸¦ ¿¹¹æÇϱâ À§ÇÑ ½Ã½ºÅÛÀ» Á¦¾ÈÇÑ´Ù. À̸¦ À§ÇÏ¿© µö·¯´× ±â¼úÀ» È°¿ëÇÏ¿© Ⱦ´Üº¸µµ¿¡ Á¢±ÙÇÏ´Â Â÷·®ÀÇ À§Ä¡ º¯È¸¦ ¸ð´ÏÅ͸µÇϸç ÃÖÀûÈµÈ ºÐ·ù ¸ðµ¨°ú ÆĶó¹ÌÅ͸¦ »ç¿ëÇÏ¿© Â÷·® ¼Óµµ¿¡ µû¸¥ À§Çè Á¤µµ¸¦ ¿¹ÃøÇÑ´Ù. ½ÇÁ¦ Â÷·® ÁÖÇà ¿µ»óÀ» ÀÌ¿ëÇÑ ½Ã¹Ä·¹ÀÌ¼Ç ¹× ½ÇÇèÀ» ¹ÙÅÁÀ¸·Î Á¦¾ÈÇÏ´Â ½Ã½ºÅÛÀÇ ±³Åë»ç°í ¿¹¹æ È°¿ë °¡´É¼ºÀ» È®ÀÎÇÏ¿´´Ù. |
¿µ¹®³»¿ë (English Abstract) |
As the individual vehicle penetration rate of Korea continues to increase, the rate of traffic accidents tend to increase as well. In particular, among the pedestrian accidents a high ratio of accidents of children and the elderly at crosswalks are observed; therefore, a lot of studies and efforts have been made to prevent crosswalk traffic accidents. This paper proposes a system to prevent pedestrian traffic accidents by providing adaptive notification through speed monitoring of vehicles approaching the crosswalk. To this end, a deep learning technology is used to monitor changes in the position of the vehicle approaching the crosswalk, and well-optimized classification models and parameters are used to estimate the speed of a vehicle, thereby predicting the degree of safety around the crosswalk. Through the experiments and simulations using real vehicle driving images, we show the feasibility of the proposed system to prevent traffic accidents. |
Å°¿öµå(Keyword) |
Deep learning
Crosswalk
Vehicle speed
Adaptive alarm
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|