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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹ÀÎÅͳÝÁ¤º¸ÇÐȸ ³í¹®Áö

Çѱ¹ÀÎÅͳÝÁ¤º¸ÇÐȸ ³í¹®Áö

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) °í¼Ó ÇØ»ó °´Ã¼ ºÐ·ù¸¦ À§ÇÑ ¾çÀÚÈ­ Àû¿ë ±â¹Ý CNN µö·¯´× ¸ðµ¨ ¼º´É ºñ±³ ºÐ¼®
¿µ¹®Á¦¸ñ(English Title) Comparative Analysis of CNN Deep Learning Model Performance Based on Quantification Application for High-Speed Marine Object Classification
ÀúÀÚ(Author) À̼ºÁÖ   ÀÌÈ¿Âù   ¼ÛÇöÇР  ÀüÈ£¼®   ÀÓÅÂÈ£   Seong-Ju Lee   Hyo-Chan Lee   Hyun-Hak Song   Ho-Seok Jeon   Tae-ho Im  
¿ø¹®¼ö·Ïó(Citation) VOL 22 NO. 02 PP. 0059 ~ 0068 (2021. 04)
Çѱ۳»¿ë
(Korean Abstract)
ÃÖ±Ù ±Þ¼Óµµ·Î ¼ºÀåÇÏ°í ÀÖ´Â ÀΰøÁö´É ±â¼úÀÌ ÀÚÀ²¿îÇ×¼±¹Ú°ú °°Àº ÇØ»ó ȯ°æ¿¡¼­µµ Àû¿ëµÇ±â ½ÃÀÛÇϸ鼭 µðÁöÅÐ ¿µ»ó¿¡ ƯȭµÈ CNN ±â¹ÝÀÇ ¸ðµ¨À» Àû¿ëÇÏ´Â °ü·Ã ¿¬±¸°¡ È°¹ßÈ÷ ÁøÇàµÇ°í ÀÖ´Ù. ÀÌ·¯ÇÑ ÇØ»ó ¼­ºñ½ºÀÇ °æ¿ì ÀÎÀû °ú½ÇÀ» ÁÙÀ̱â À§ÇØ Ãæµ¹ À§ÇèÀÌ ÀÖ´Â ºÎÀ¯¹°À» °¨ÁöÇϰųª ¼±¹Ú ³»ºÎÀÇ È­Àç µî ¿©·¯ °¡Áö ±â¼úÀÌ Á¢¸ñµÇ±â¿¡ ½Ç½Ã°£ 󸮰¡ ¸Å¿ì Áß¿äÇÏ´Ù. ±×·¯³ª ±â´ÉÀÌ Ãß°¡µÉ¼ö·Ï ÇÁ·Î¼¼¼­ÀÇ Á¦Ç° °¡°ÝÀÌ Áõ°¡ÇÏ´Â ¹®Á¦°¡ Á¸ÀçÇØ ¼ÒÇü ¼±¹ÚÀÇ ¼±Áֵ鿡°Ô´Â ºñ¿ëÀûÀÎ Ãø¸é¿¡¼­ ºÎ´ãÀÌ µÈ´Ù. ¶ÇÇÑ ´ëÇü ¼±¹ÚÀÇ °æ¿ì ÀÚÀ²¿îÇ×¼±¹ÚÀÇ ½Ã½ºÅÛÀ» °¨¾ÈÇÒ ¶§, ¿¬»ê ¼ÓµµÀÇ ¼º´É Çâ»óÀ» À§ÇØ º¹Àâµµ°¡ ³ôÀº µö·¯´× ¸ðµ¨ÀÇ ¼º´ÉÀ» °³¼±ÇÏ´Â ¹æ¹ýÀÌ ÇÊ¿äÇÏ´Ù. µû¶ó¼­ º» ³í¹®¿¡¼­´Â µö·¯´× ¸ðµ¨¿¡ °æ·®È­ ±â¹ýÀ» Àû¿ëÇØ Á¤È®µµ¸¦ À¯ÁöÇϸ鼭 °í¼ÓÀ¸·Î ó¸®ÇÒ ¼ö ÀÖ´Â ¹æ¹ý¿¡ ´ëÇØÁ¦¾ÈÇÑ´Ù. ¸ÕÀú ÇØ»ó ºÎÀ¯¹° °ËÃâ¿¡ ÀûÇÕÇÑ ¿µ»ó Àü󸮸¦ ÁøÇàÇÏ¿© È¿À²ÀûÀ¸·Î CNN ±â¹Ý ½Å°æ¸Á ¸ðµ¨ ÀԷ¿¡ ¿µ»ó µ¥ÀÌÅÍ°¡ Àü´ÞµÉ ¼ö ÀÖµµ·Ï ÇÏ¿´´Ù. ¶ÇÇÑ, ½Å°æ¸Á ¸ðµ¨ÀÇ ¾Ë°í¸®Áò °æ·®È­ ±â¹ý Áß ÇϳªÀÎ ÇнÀ ÈÄ ÆĶó¹ÌÅÍ ¾çÀÚÈ­ ±â¹ýÀ» Àû¿ëÇÏ¿© ¸ðµ¨ÀÇ ¸Þ¸ð¸® ¿ë·®À» ÁÙÀ̸鼭 Ãß·Ð ºÎºÐÀÇ Ã³¸® ¼Óµµ¸¦ Áõ°¡½ÃÄ×´Ù. ¾çÀÚÈ­ ±â¹ýÀÌ Àû¿ëµÈ ¸ðµ¨À» ÀúÀü·Â ÀÓº£µðµå º¸µå¿¡ Àû¿ë½ÃÄÑ Á¤È®µµ¿Í ó¸® ¼Óµµ¸¦ »ç¿ëÇÏ´Â ÀÓº£µðµå ¼º´ÉÀ» °í·ÁÇÏ¿© ¼³°èÇÏ´Â ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. Á¦¾ÈÇÏ´Â ¹æ¹ý Áß Á¤È®µµ ¼Õ½ÇÀÌ Á¦ÀÏ ÃÖ¼ÒÈ­µÇ´Â ¸ðµ¨À» È°¿ëÇØ ÀúÀü·Â ÀÓº£µðµå º¸µå¿¡ ºñ±³ÇÏ¿© ±âÁ¸º¸´Ù ÃÖ´ë 4~5¹è ó¸® ¼Óµµ¸¦ °³¼±ÇÒ ¼ö ÀÖ¾ú´Ù.
¿µ¹®³»¿ë
(English Abstract)
As artificial intelligence(AI) technologies, which have made rapid growth recently, began to be applied to the marine environment such as ships, there have been active researches on the application of CNN-based models specialized for digital videos. In E-Navigation service, which is combined with various technologies to detect floating objects of clash risk to reduce human errors and prevent fires inside ships, real-time processing is of huge importance. More functions added, however, mean a need for high-performance processes, which raises prices and poses a cost burden on shipowners. This study thus set out to propose a method capable of processing information at a high rate while maintaining the accuracy by applying Quantization techniques of a deep learning model. First, videos were pre-processed fit for the detection of floating matters in the sea to ensure the efficient transmission of video data to the deep learning entry. Secondly, the quantization technique, one of lightweight techniques for a deep learning model, was applied to reduce the usage rate of memory and increase the processing speed. Finally, the proposed deep learning model to which video pre-processing and quantization were applied was applied to various embedded boards to measure its accuracy and processing speed and test its performance. The proposed method was able to reduce the usage of memory capacity four times and improve the processing speed about four to five times while maintaining the old accuracy of recognition.
Å°¿öµå(Keyword) CNN   ¸ðµ¨ ¾çÀÚÈ­   ¿µ»ó 󸮠  ¼±¹Ú °´Ã¼ ºÐ·ù   CNN   Model Quantization   Image pre-processing   Ship classification  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå