Ç¥ÁØÈ­ Âü¿©¾È³»

TTAÀÇ Ç¥ÁØÇöȲ

Ȩ > Ç¥ÁØÈ­ °³¿ä > TTAÀÇ Ç¥ÁØÇöȲ

Ç¥ÁعøÈ£ TTAK.KO-10.1310 ±¸Ç¥ÁعøÈ£
Á¦°³Á¤ÀÏ 2021-12-08 ÃÑÆäÀÌÁö 5
ÇѱÛÇ¥ÁØ¸í ¼­Æ÷Æ® º¤ÅÍ ¸Ó½Å(SVM) ºÐ·ù±â¸¦ »ç¿ëÇÏ´Â Á¤±³ÇÑ µ¥ÀÌÅÍ Àü¼Û LED °¨Áö ¹æ½Ä
¿µ¹®Ç¥Áظí Accurate data transmitting LED detection scheme employing support vector machine (SVM) classifier
Çѱ۳»¿ë¿ä¾à ±âÁ¸ ±¤ Ä«¸Þ¶ó Åë½Å(OCC) ½Ã½ºÅÛ¿¡¼­´Â ¼Û½Å LED°¡ Ä«¸Þ¶ó ¼ö½Å±â¿¡ ÀÇÇØ °¨ÁöµÈ´Ù. ±×·¯³ª V2X ³ª ·Îº¿ °øÇаú °°Àº ½Ç½Ã°£ ¾ÖÇø®ÄÉÀ̼ǿ¡¼­´Â ½ÇÁ¦ ¼Û½Å±â°¡ ¾Æ´Ï±â ¶§¹®¿¡ LED°¡ ¸¹ÀÌ ÀÖÀ» ¼ö ÀÖ´Ù. ÀÌ °æ¿ì µðÄÚµùµÈ µ¥ÀÌÅÍ¿¡¼­ ½ºÇ»¸®¾î½º µ¥ÀÌÅÍ ºñÆ®ÀÇ È®·üÀÌ ³ô¾ÆÁ® ºñÆ®¿À·ùÀ²(BER)ÀÌ Ä¿Áú °ÍÀ¸·Î ¿¹»óµÈ´Ù. ´õ¿íÀÌ Ä¸Ã³µÈ LED ¿µ¿ªÀ» ´Ù¸¥ ºñ Àü¼Û LED ¿µ¿ª°ú ÇÔ²² ó¸®ÇÏ·Á¸é ¼ö½Å±âÀÇ º¹À⼺ÀÌ Áõ°¡ÇÑ´Ù. ÀÌ ¹®Á¦¸¦ Áö´ÉÀûÀ¸·Î ó¸®Çϱâ À§ÇØ ¾ÈÁ¤ÀûÀÎ OCC ½Ã½ºÅÛÀ» À§ÇÑ ±âÁ¸ µ¥ÀÌÅÍ µðÄÚµù ¹æ½Ä¿¡¼­ SVM°ú CNNÀ» °áÇÕÇÏ´Â ¹æ½ÄÀ» Á¦¾ÈÇÑ´Ù.
¿µ¹®³»¿ë¿ä¾à In traditional OCC systems, the transmit LED is detected by the camera receiver. However, in real-time applications such as V2X or robotics, there may be many LEDs that are not real transmitters. In this case, a higher probability of spurious data bits in the decoded data is expected, resulting in a higher BER. Moreover, processing the captured LED area along with other non-transmitting LED areas increases the complexity of the receiver. To intelligently deal with this problem, we propose a method combining SVM and CNN in the existing data decoding method for a stable OCC system.
±¹Á¦Ç¥ÁØ
°ü·ÃÆÄÀÏ TTAK.KO-10.1310_[1].pdf TTAK.KO-10.1310_[1].pdf            

ÀÌÀü
°¡º¯ÆÞ½ºÆø º¯Á¶±â¹Ý ±¤ÇÐ Ä«¸Þ¶ó Åë½ÅÀ» À§ÇÑ ½Å°æ¸Á ¾ÆÅ°ÅØó
´ÙÀ½
½Ç³»¿ë ½º¸¶Æ®Àç¹è½Ã½ºÅÛ ±â´É±¸Á¶ ¹× ÀÎÅÍÆäÀ̽º ¿ä±¸»çÇ×