• 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¸Åë½ÅÇÐȸ ³í¹®Áö (Journal of the Korea Institute of Information and Communication Engineering)

Çѱ¹Á¤º¸Åë½ÅÇÐȸ ³í¹®Áö (Journal of the Korea Institute of Information and Communication Engineering)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) DBNÀ» ÀÌ¿ëÇÑ ´ÙÁß ¹æÀ§ µ¥ÀÌÅÍ ±â¹Ý ´Éµ¿¼Ò³ª Ç¥Àû ½Äº°
¿µ¹®Á¦¸ñ(English Title) Multiaspect-based Active Sonar Target Classification Using Deep Belief Network
ÀúÀÚ(Author) ±èµ¿¿í   ¹è°Ç¼º   ¼®Á¾¿ø   Dong-wook Kim   Keun-sung Bae   Jong-won Seok  
¿ø¹®¼ö·Ïó(Citation) VOL 22 NO. 03 PP. 0418 ~ 0424 (2018. 03)
Çѱ۳»¿ë
(Korean Abstract)
¼öÁß Ç¥Àû ŽÁö ¹× ½Äº°Àº ±º»ç ¹× ºñ±º»çÀûÀ¸·Î Áß¿äÇÑ ¹®Á¦ÀÌ´Ù. ÃÖ±Ù ÆÐÅÏÀÎ½Ä ºÐ¾ß¿¡¼­ µö·¯´× ±â¼úÀÌ ¹ßÀüµÇ¸é¼­ ¸¹Àº ¼º´É°³¼± °á°ú°¡ ¹ßÇ¥µÇ°í ÀÖ´Ù. ±×Áß DBN(Deep Belief Network)±â¹ýÀº DNN(Deep Neural Network)À» »çÀü ÈÆ·ÃÇϴµ¥ »ç¿ëµÇ¾î ÁÁÀº ¼º´ÉÀ» º¸¿©ÁÖ°í ÀÖ´Ù. º» ³í¹®¿¡¼­´Â ´Éµ¿ ¼Ò³ª¸¦ ÀÌ¿ëÇÑ ¼öÁß Ç¥ÀûÀÇ ½Äº° ¹®Á¦¿¡ DBNÀ» »ç¿ëÇÏ¿© ½ÇÇèÀ» ÁøÇàÇÏ°í, ±× °á°ú¸¦ ºñ±³ÇÏ¿´´Ù. Ç¥Àû½ÅÈ£´Â 3Â÷¿ø ÇÏÀ̶óÀÌÆ® ¸ðµ¨À» »ç¿ëÇÏ¿© ÇÕ¼ºµÈ ´Éµ¿ ¼Ò³ª ½ÅÈ£¸¦ »ç¿ëÇÏ¿´°í, Ư¡ÃßÃâ ¹æ¹ýÀ¸·Î´Â FrFT(Fractional Fourier Transform) ±â¹ÝÀÇ Æ¯Â¡ÃßÃâÀ» »ç¿ëÇÏ¿´´Ù. ´ÜÀÏ ¼¾¼­, Áï, ´ÜÀÏ ¹æÀ§ µ¥ÀÌÅÍ ±â¹ÝÀÇ ½ÇÇè¿¡¼­ DBNÀ» ÀÌ¿ëÇÑ ½Äº° °á°ú´Â ±âÁ¸ÀÇ BPNN(Back Propagation Neural Network)¿¡ ºñÇØ ¾à 3.83 % Çâ»óµÇ¾ú´Ù. ¶ÇÇÑ, ´ÙÁß ¹æÀ§ ±â¹ÝÀÇ ½Äº° ½ÇÇè¿¡¼­´Â °üÃø¿­ÀÇ °³¼ö°¡ 3À» ÃÊ°úÇϸé 95% ÀÌ»óÀÇ ¼º´ÉÀ» ¾òÀ» ¼ö ÀÖ¾ú´Ù.
¿µ¹®³»¿ë
(English Abstract)
Detection and classification of underwater targets is an important issue for both military and non-military purposes. Recently, many performance improvements are being reported in the field of pattern recognition with the development of deep learning technology. Among the results, DBN showed good performance when used for pre-training of DNN. In this paper, DBN was used for the classification of underwater targets using active sonar, and the results are compared with that of the conventional BPNN. We synthesized active sonar target signals using 3-dimensional highlight model. Then, features were extracted based on FrFT. In the single aspect based experiment, the classification result using DBN was improved about 3.83% compared with the BPNN. In the case of multi-aspect based experiment, a performance of 95% or more is obtained when the number of observation sequence exceeds three.
Å°¿öµå(Keyword) ´Éµ¿ ¼Ò³ª   ÇÏÀ̶óÀÌÆ® ¸ðµ¨   ½ÉÃþÇнÀ   ½ÉÃþ ½Å·Ú¸Á   Fractional Ǫ¸®¿¡ º¯È¯   Active Sonar   Highlight model   Deep learning   Deep belief network   Fractional fourier transform  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå