Processing math: 100%
  • ๋Œ€ํ•œ์ „๊ธฐํ•™ํšŒ
Mobile QR Code QR CODE : The Transactions of the Korean Institute of Electrical Engineers
  • COPE
  • kcse
  • ํ•œ๊ตญ๊ณผํ•™๊ธฐ์ˆ ๋‹จ์ฒด์ด์—ฐํ•ฉํšŒ
  • ํ•œ๊ตญํ•™์ˆ ์ง€์ธ์šฉ์ƒ‰์ธ
  • Scopus
  • crossref
  • orcid

  1. (Dept. of Robotics and Automation Engineering, Hoseo University, Korea.)
  2. (Dept. of Robotics and Automation Engineering, Hoseo University, Korea.)



Automotive mobile robot, Computer vision, Image processing, Grid tracking robot

1. ์„œ ๋ก 

์ตœ๊ทผ ๋กœ๋ด‡์˜ ์œ„์น˜ ์ธ์‹ ๊ธฐ์ˆ ์— ๊ธฐ๋ฐ˜์„ ๋‘” ์‹ค๋‚ด ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡ ๊ธฐ์ˆ ์ด ์ฃผ๋ชฉ์„ ๋ฐ›๊ณ  ์žˆ๋‹ค(1-3). ๋กœ๋ด‡์˜ ์œ„์น˜ ์ธ์‹์€ ์ด๋™ ๋กœ๋ด‡์˜ ์œ„์น˜์™€ ๋ฐฉํ–ฅ์„ ์ธ์‹ํ•˜๊ฑฐ๋‚˜ ํŒŒ์•…ํ•˜๋Š” ๊ธฐ์ˆ ์ด๋‹ค(1). ์ด๋™ ๋กœ๋ด‡์˜ ํ•ต์‹ฌ ๊ธฐ์ˆ ์ด๋ผ ํ•  ์ˆ˜ ์žˆ๋Š” ์œ„์น˜ ์ธ์‹์€ ๋กœ๋ด‡์ด ์†ํ•œ ๊ณต๊ฐ„์—์„œ 2์ฐจ์› ์ขŒํ‘œ์™€ ๋กœ๋ด‡์ด ํ–ฅํ•œ ๋ฐฉํ–ฅ ์ •๋ณด๋ฅผ ํŒŒ์•…ํ•˜๋Š” ๊ฒƒ์ด๋‹ค(2). ๊ทธ๋™์•ˆ ์œ„์น˜ ์ธ์‹ ์—ฐ๊ตฌ๋Š” ์ฃผ๋กœ ์—”์ฝ”๋”(Encoder), ์ ์™ธ์„ , ์ดˆ์ŒํŒŒ ๋“ฑ์„ ์žฅ์ฐฉํ•˜์—ฌ ์˜ค๋„๋ฉ”ํƒ€(Odometer) ์ •๋ณด์— ์˜์กดํ•ด ๋กœ๋ด‡์˜ ์ž๊ธฐ ์œ„์น˜ ์ง€ํ‘œ๋ฅผ ์ธ์‹ํ•˜๊ณ , ์žฅ์• ๋ฌผ์˜ ์—ฌ๋ถ€๋ฅผ ์ธ์‹ํ•˜์˜€๋‹ค(3).

์ตœ๊ทผ์—๋Š” ์ด๋™ ๋กœ๋ด‡์ด ๋ฏธ์ง€์˜ ์˜์—ญ์„ ์ฃผํ–‰ํ•˜๋ฉด์„œ ์ฃผ๋ณ€ ํ™˜๊ฒฝ์„ ์ธ์ง€ํ•˜๊ณ  ์ง€๋„๋ฅผ ์ž‘์„ฑํ•˜๋ฉด์„œ ๋™์‹œ์— ๊ทธ ํ™˜๊ฒฝ์— ๋Œ€ํ•œ ๋กœ๋ด‡์˜ ์ƒ๋Œ€์ ์ธ ์œ„์น˜๋ฅผ ์ถ”์ •ํ•˜๋Š” SLAM(Simultaneous Localization And Mapping)์ด ๋งŽ์ด ํ™œ์šฉ๋˜๊ณ  ์žˆ๋‹ค(4). ๋กœ๋ด‡์˜ ์ด๋™ ํ™˜๊ฒฝ์— ๋Œ€ํ•œ ์ •๋ณด๋Š” ์ฃผ๋กœ ๋กœ๋ด‡์— ๋ถ€์ฐฉ๋œ ๊ฑฐ๋ฆฌ ์„ผ์„œ(LiDAR, Laser, ์ดˆ์ŒํŒŒ, ์ ์™ธ์„  ๋“ฑ)๋ฅผ ์ด์šฉํ•˜์—ฌ ํš๋“ํ•˜๋Š” ๋ฐฉ๋ฒ•๊ณผ ๋น„์ „(Vision) ์„ผ์„œ๋ฅผ ์ด์šฉํ•˜์—ฌ ์˜์ƒ ์ฒ˜๋ฆฌ ๊ณผ์ •์„ ํ†ตํ•˜์—ฌ ์ •๋ณด๋ฅผ ํš๋“ํ•˜๋Š” ๋ฐฉ๋ฒ•์ด ์žˆ๋‹ค(5-7).

ํ˜„์žฌ๊นŒ์ง€ ์—ฐ๊ตฌ๋œ ๋‹ค์–‘ํ•œ ๋กœ๋ด‡์˜ ์œ„์น˜ ์ธ์‹๊ณผ ๋„ค๋น„๊ฒŒ์ด์…˜ ๊ธฐ์ˆ ์€ ์ธ๊ณต์ง€ํ‘œ๋‚˜ ๋น„์ „์„ ๊ธฐ๋ฐ˜์œผ๋กœ ์ง„ํ–‰ํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ์ด์™€ ๊ฐ™์€ ๊ธฐ์ˆ ์€ ์œ„์น˜ ์ธ์‹์˜ ์˜ค์ฐจ๋ฒ”์œ„๊ฐ€ ์ผ์ • ๊ธฐ์ค€ ์ด์ƒ ๋ฐœ์ƒํ•˜๊ณ , ๊ธฐ์ดˆ ์‹œ๊ณต๋น„์šฉ์ด ๋†’์•„ ๋ณด๋‹ค ์ •ํ™•ํ•œ ์œ„์น˜ ์ธ์‹๊ณผ ๋‚ฎ์€ ์‹œ๊ณต๋น„์šฉ์ด ์š”๊ตฌ๋˜๋Š” ๊ธฐ์ˆ ์„ ํ•„์š”๋กœ ํ•œ๋‹ค(1).

์ธ๊ฐ„์˜ ์‹œ๊ฐ ์ง€๋Šฅ์€ ์‹œ๊ฐ์  ํ™˜๊ฒฝ์„ ์ž์‹ ์˜ ์ƒ์กด์— ๋งž๊ฒŒ ํ™œ์šฉํ•˜๋Š” ์—ญ๋Ÿ‰์„ ์˜๋ฏธํ•œ๋‹ค. ์ธ๊ฐ„์€ ์‹œ๊ฐ ์ง€๋Šฅ์„ ์ด์šฉํ•˜์—ฌ ๋ฏธ์ง€์˜ ์‹ค๋‚ด ๊ณต๊ฐ„์—์„œ ๋‹ค๋ฅธ ๊ฐ๊ฐ์˜ ๋„์›€ ์—†์ด ์‹œ๊ฐ ์ธ์ง€ ์ •๋ณด๋งŒ์œผ๋กœ๋„ ๋ชฉํ‘œ ์œ„์น˜๋ฅผ ์ฐพ์•„๊ฐˆ ์ˆ˜ ์žˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด ๋ฐฐ๋‹ฌ ๋ผ์ด๋”๊ฐ€ ์ฒ˜์Œ ๋ฐฉ๋ฌธํ•˜๋Š” ๊ฑด๋ฌผ์˜ ํ˜ธ์‹ค ์ •๋ณด๋งŒ ์•Œ๊ณ  ์žˆ๋‹ค๊ณ  ํ•˜๋”๋ผ๋„ ๋ผ์ด๋”๋Š” ๋‹ค๋ฅธ ๊ฐ๊ฐ์— ์˜์กด ํ•˜์ง€ ์•Š๊ณ  ์˜ค์ง ์‹œ๊ฐ ์ •๋ณด๋งŒ์„ ์ด์šฉํ•˜์—ฌ ๋ฐฐ๋‹ฌ ์žฅ์†Œ๋ฅผ ์ฐพ์„ ์ˆ˜ ์žˆ๋‹ค. ์ด์™€ ๊ฐ™์ด ์ธ๊ฐ„์˜ ์‹œ๊ฐ ์ง€๋Šฅ์„ ์ด์šฉํ•˜๋ฉด ์‹ค๋‚ด ๊ณต๊ฐ„์— ์„ค์น˜๋œ ์ตœ์†Œํ•œ์˜ ์‹œ๊ฐ ์ •๋ณด๋งŒ์œผ๋กœ๋„ ๋กœ๋ด‡์˜ ์‹ค๋‚ด ์ž์œจ์ฃผํ–‰์ด ๊ฐ€๋Šฅํ•˜๊ฒŒ ๋  ๊ฒƒ์ด๋‹ค.

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ์ธ๊ฐ„์˜ ์‹œ๊ฐ ์ง€๋Šฅ์„ ๋ชจ๋ฐฉํ•˜์—ฌ ๋กœ๋ด‡์ด ์‹ค๋‚ด ๊ณต๊ฐ„์˜ ๋ฐ”๋‹ฅ ๋ฉด์— ์„ค์น˜๋œ ํƒ€์ผ์˜ ๊ฒฉ์ž๋ฌด๋Šฌ ์ •๋ณด๋ฅผ ์ด์šฉํ•˜์—ฌ ์œ„์น˜๋ฅผ ์ธ์‹ํ•˜๊ณ  ์ฃผํ–‰ ๋ฐฉํ–ฅ์„ ๊ฒฐ์ •ํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์ œ์•ˆํ•œ๋‹ค. ์ด์™€ ๊ฐ™์€ ๋กœ๋ด‡์˜ ์‹ค๋‚ด ์ฃผํ–‰ ๋ฐฉ๋ฒ•์€ ์‹ค๋‚ด ๋กœ๋ด‡ ์‹œ์Šคํ…œ์„ ์œ„ํ•œ ๋ณ„๋„์˜ ์‹œ๊ณต์ด๋‚˜ ์žฅ์น˜์˜ ์ถ”๊ฐ€ ์—†์ด ๋‹จ์ผ ์นด๋ฉ”๋ผ์˜ ์ •๋ณด๋งŒ์œผ๋กœ ์ด๋™ ๋กœ๋ด‡์˜ ์œ„์น˜ ์ธ์‹๊ณผ ์ฃผํ–‰ ์ œ์–ด๋ฅผ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ•˜๋ฏ€๋กœ ๊ธฐ์กด์˜ ๋ฐฉ๋ฒ•์— ๋น„ํ•ด ๊ฐ„๋‹จํ•˜๊ณ  ๊ฒฝ์ œ์ ์ด๋‹ค.

2. ์ด๋™ ๋กœ๋ด‡ ์‹œ์Šคํ…œ ์„ค๊ณ„

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ์‹ค๋‚ด ๋ฐ”๋‹ฅ ํƒ€์ผ์˜ ๊ฒฉ์ž๋ฌด๋Šฌ๋ฅผ ์ด์šฉํ•˜์—ฌ ์ธ๊ฐ„์˜ ์‹œ๊ฐ ์ง€๋Šฅ์„ ๋ชจ๋ฐฉํ•œ ์‹ค๋‚ด ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์„ ๊ตฌํ˜„ํ•˜์˜€๋‹ค.

๋กœ๋ด‡์˜ ๊ตฌ๋™์„ ์œ„ํ•˜์—ฌ 2๊ฐœ์˜ BLDC ๋ชจํ„ฐ๋ฅผ ์ด์šฉํ•œ ์ฐจ๋™๊ตฌ๋™๋ฐฉ์‹์„ ์ ์šฉํ•˜์˜€๋‹ค. ๋กœ๋ด‡์˜ ํ•˜๋‹จ ์ค‘์•™ ์–‘์ชฝ์—๋Š” 2๊ฐœ์˜ BLDC ๋ชจํ„ฐ๋ฅผ ์žฅ์ฐฉํ•˜์˜€๊ณ , ์ „๋ฉด๊ณผ ํ›„๋ฉด์—๋Š” ๊ตฌ๋™๋ถ€๊ฐ€ ์—†๋Š” 4๊ฐœ์˜ ๋ณด์กฐ ๋ฐ”ํ€ด๋ฅผ ๊ตฌ์„ฑํ•˜์˜€๋‹ค. ์ „๋ฉด ์ƒ๋‹จ์—๋Š” ๋ฐ”๋‹ฅ ๋ฉด์˜ ํƒ€์ผ์„ ์ดฌ์˜ํ•˜๊ธฐ ์œ„ํ•ด CMOS ์นด๋ฉ”๋ผ๋ฅผ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, ์ฃผํ–‰ ๋ฐฉํ–ฅ์˜ ์žฅ์• ๋ฌผ์„ ์ธ์‹ํ•˜๊ธฐ ์œ„ํ•ด ์ดˆ์ŒํŒŒ ์„ผ์„œ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ์•„๋ž˜ ๊ทธ๋ฆผ 1์€ ๋ณธ ๋…ผ๋ฌธ์—์„œ ์‹คํ—˜์„ ์œ„ํ•ด ์ œ์ž‘ํ•œ ์‹ค๋‚ด ์ž์œจ ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์™ธํ˜•์ด๋‹ค.

๊ทธ๋ฆผ. 1. ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡

Fig. 1. Autonomous mobile robot that tracks tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig1.png

2.1 ์ด๋™ ๋กœ๋ด‡ ์‹œ์Šคํ…œ ๊ตฌ์„ฑ

๊ทธ๋ฆผ 2๋Š” ๋ณธ ๋…ผ๋ฌธ์—์„œ ๊ตฌํ˜„ํ•œ ์‹ค๋‚ด ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡ ์‹œ์Šคํ…œ์˜ ๊ธฐ๋Šฅ ๋ธ”๋ก๋„์ด๋‹ค. ์ œ์–ด๋ถ€๋Š” Microchip์‚ฌ์˜ PIC16F884๊ธฐ๋ฐ˜ ์ž„๋ฒ ๋””๋“œ ๋ณด๋“œ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ์˜์ƒ ์ดฌ์˜๊ณผ ์ปดํ“จํ„ฐ ๋น„์ „ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์ฒ˜๋ฆฌ, ๊ทธ๋ฆฌ๊ณ  ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์ œ์–ด๋ฅผ ์œ„ํ•ด ๋…ธํŠธ๋ถ์„ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, ์ด๋™ ๋กœ๋ด‡์˜ ์‹ค์‹œ๊ฐ„ ์ฃผํ–‰ ๋ชจ๋‹ˆํ„ฐ๋ง๊ณผ ์›๊ฒฉ ์ œ์–ด๋ฅผ ์œ„ํ•ด ๋ณ„๋„์˜ PC๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค.

๊ทธ๋ฆผ. 2. ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡ ์‹œ์Šคํ…œ ๊ธฐ๋Šฅ ๋ธ”๋ก๋„

Fig. 2. System function block diagram of an autonomous mobile robot that tracks tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig2.png

2.2 ์ด๋™ ๋กœ๋ด‡ ์‹œ์Šคํ…œ ์ œ์–ด๊ธฐ ์„ค๊ณ„

์ด๋™ ๋กœ๋ด‡ ์‹œ์Šคํ…œ ์ œ์–ด๊ธฐ์˜ ํ•˜๋“œ์›จ์–ด ๊ตฌ์„ฑ์€ ๊ทธ๋ฆผ 3๊ณผ ๊ฐ™๋‹ค. CPU๋Š” Microchip์‚ฌ์˜ PIC16F884 8bit ๋งˆ์ดํฌ๋กœ์ปจํŠธ๋กค๋Ÿฌ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ๋กœ๋ด‡์„ ๊ตฌ๋™ํ•˜๊ธฐ ์œ„ํ•œ ๋ชจํ„ฐ๋Š” 2๊ฐœ์˜ In-Wheel BLDC ๋ชจํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ์žฅ์• ๋ฌผ ๊ฐ์ง€๋ฅผ ์œ„ํ•˜์—ฌ ์ „๋ฐฉ๊ณผ ํ›„๋ฐฉ์— ๊ฐ๊ฐ 1๊ฐœ์˜ ์ดˆ์ŒํŒŒ ์„ผ์„œ๋ฅผ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, ์˜์ƒํš๋“์„ ์œ„ํ•˜์—ฌ ๋…ธํŠธ๋ถ๊ณผ ์›น์บ (Webcam)์„ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ํš๋“ํ•œ ์˜์ƒ์€ ์œ„์น˜ ์ถ”์ • ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ด์šฉํ•ด ์ด๋™ ๋กœ๋ด‡์˜ ํ˜„ ์œ„์น˜๋ฅผ ์ถ”์ •ํ•˜๋Š”๋ฐ ์‚ฌ์šฉ๋˜๊ณ , ๊ณ„์‚ฐ๋œ ํ˜„ ์œ„์น˜๋Š” ์ฃผํ–‰ ์ œ์–ด ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ด์šฉํ•˜์—ฌ CPU์— ์ฐจ๋™ ์ฃผํ–‰ ์‹ ํ˜ธ๋ฅผ ์ „์†กํ•˜๊ธฐ ์œ„ํ•ด ์‚ฌ์šฉ๋œ๋‹ค. ์ด๋ฅผ ํ†ตํ•ด ์‹ค๋‚ด ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ๋ฐฉํ–ฅ๊ณผ ์†๋„๋ฅผ ์ œ์–ดํ•  ์ˆ˜ ์žˆ๋‹ค.

๊ทธ๋ฆผ. 3. ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡ ์‹œ์Šคํ…œ ์ œ์–ด๊ธฐ์˜ ํ•˜๋“œ์›จ์–ด ๊ตฌ์„ฑ

Fig. 3. Hardware configuration for system controller of an autonomous mobile robot tracking tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig3.png

3. ์ด๋™ ๋กœ๋ด‡์˜ ์œ„์น˜ ์ถ”์ •

3.1 ๊ฐ€์ด๋“œ๋ผ์ธ ์ถ”์ถœ์„ ์œ„ํ•œ ์ด๋ฏธ์ง€ ์ „์ฒ˜๋ฆฌ

์ด๋™ ๋กœ๋ด‡์˜ ํ˜„์žฌ ์œ„์น˜์™€ ์ง„ํ–‰ ๋ฐฉํ–ฅ์˜ ์ฃผํ–‰ ์ •๋ณด๋ฅผ ํš๋“ํ•˜๊ธฐ ์œ„ํ•ด์„œ ๋ฐ”๋‹ฅ ํƒ€์ผ์˜ ์˜์ƒ์œผ๋กœ๋ถ€ํ„ฐ ๊ฐ€์ด๋“œ๋ผ์ธ(Guideline)์„ ์ถ”์ถœํ•œ๋‹ค. ๊ทธ๋ฆผ 4๋Š” ์นด๋ฉ”๋ผ๋ฅผ ํ†ตํ•ด ํš๋“ํ•œ ๋ฐ”๋‹ฅ ํƒ€์ผ์˜ ์›๋ณธ ์ด๋ฏธ์ง€์ด๋‹ค.

๊ทธ๋ฆผ. 4. ๋ฐ”๋‹ฅ ํƒ€์ผ ์›๋ณธ ์ด๋ฏธ์ง€

Fig. 4. Original images of tiles installed on the floor

../../Resources/kiee/KIEE.2021.70.6.945/fig4.png

๊ทธ๋ฆผ 5๋Š” ๋ฐ”๋‹ฅ ํƒ€์ผ ์›๋ณธ ์˜์ƒ์œผ๋กœ๋ถ€ํ„ฐ ํƒ€์ผ์˜ ๊ฒฉ์ž๋ฌด๋Šฌ ์ •๋ณด๋ฅผ ์–ป๊ณ  ์ด๋ฅผ ํ†ตํ•ด ๊ฐ€์ด๋“œ๋ผ์ธ์„ ์ถ”์ถœํ•˜๊ธฐ ์œ„ํ•œ ์ด๋ฏธ์ง€ ์ „์ฒ˜๋ฆฌ ๊ณผ์ •์„ ๋ณด์—ฌ์ฃผ๋Š” ์ˆœ์„œ๋„์ด๋‹ค.

3.2๋ถ€ํ„ฐ 3.6๊นŒ์ง€๋Š” ์ด๋™ ๋กœ๋ด‡์˜ ์œ„์น˜ ์ธ์‹๊ณผ ์ฃผํ–‰ ์ œ์–ด์— ์‚ฌ์šฉ๋  ๊ฐ€์ด๋“œ๋ผ์ธ ์ถ”์ถœ์„ ์œ„ํ•œ ์ด๋ฏธ์ง€ ์ „์ฒ˜๋ฆฌ ๊ณผ์ •์„ ์ƒ์„ธํžˆ ์„ค๋ช…ํ•œ๋‹ค.

๊ทธ๋ฆผ. 5. ๊ฐ€์ด๋“œ๋ผ์ธ ์ถ”์ถœ์„ ์œ„ํ•œ ์ด๋ฏธ์ง€ ์ „์ฒ˜๋ฆฌ ์ ˆ์ฐจ

Fig. 5. Image pre-processing procedure for guideline extraction

../../Resources/kiee/KIEE.2021.70.6.945/fig5.png

3.2 ๋…ธ์ด์ฆˆ ์ œ๊ฑฐ

์นด๋ฉ”๋ผ๋กœ๋ถ€ํ„ฐ ๋ฐ›์•„์ง„ ์˜์ƒ์€ ์กฐ๋ช… ๋“ฑ์˜ ์˜ํ–ฅ์œผ๋กœ ์ธํ•˜์—ฌ ๋งŽ์€ ๋…ธ์ด์ฆˆ๋ฅผ ํฌํ•จํ•˜๊ณ  ์žˆ๋‹ค. ํ๋ฆผํšจ๊ณผ(Blurring)์™€ ๋ชจํด๋กœ์ง€(Morphology) ์—ฐ์‚ฐ์„ ์ด์šฉํ•˜์—ฌ ์˜์ƒ์— ํฌํ•จ๋œ ๋…ธ์ด์ฆˆ๋ฅผ ์ตœ์†Œํ™” ํ•œ๋‹ค(8). ๊ทธ๋ฆผ 6์€ ์นด๋ฉ”๋ผ๋ฅผ ํ†ตํ•ด ํš๋“ํ•œ ์›๋ณธ ํƒ€์ผ ์ด๋ฏธ์ง€(๊ทธ๋ฆผ 4์˜ floor_1)์— ๊ฐ€์šฐ์‹œ์•ˆ ํ๋ฆผ ํšจ๊ณผ(Gaussian blur)์™€ ํžˆํŠธ๋ฏธ์Šค ๋ชจํด๋กœ์ง€(Hit-or-miss morphology) ์—ฐ์‚ฐ์„ ์ ์šฉํ•˜์—ฌ ์›๋ณธ ์˜์ƒ์œผ๋กœ๋ถ€ํ„ฐ ๋…ธ์ด์ฆˆ๋ฅผ ์ œ๊ฑฐํ•œ ์ด๋ฏธ์ง€์ด๋‹ค.

3.3 ์ด์ง„ํ™”

์ „์ฒ˜๋ฆฌ ๊ณผ์ •์„ ํ†ตํ•ด ๋…ธ์ด์ฆˆ๋ฅผ ์ถ•์†Œํ•œ ์ด๋ฏธ์ง€๋ฅผ ๊ทธ๋ ˆ์ด์Šค์ผ€์ผ(Grayscale)๋กœ ๋ณ€ํ™˜ํ•œ ํ›„ ์ฒ˜๋ฆฌ ์†๋„ ํ–ฅ์ƒ๊ณผ ์ž”์—ฌ ๋…ธ์ด์ฆˆ ์ œ๊ฑฐ๋ฅผ ์œ„ํ•ด ์ด์ง„ํ™”(Binarization)๋ฅผ ์ˆ˜ํ–‰ํ•œ๋‹ค.

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋ชจํด๋กœ์ง€ ์—ฐ์‚ฐ ์ˆ˜ํ–‰ ํ›„ ์ด๋ฏธ์ง€์— ๋‚จ์•„์žˆ๋Š” ์ž”์—ฌ ๋…ธ์ด์ฆˆ ์ œ๊ฑฐ๋ฅผ ์œ„ํ•ด ๊ฐ€์ค‘ํ‰๊ท (Weighted average) ์ ์‘ํ˜• ์ž„๊ณ„๊ฐ’(Adaptive threshold) ์ด์ง„ํ™” ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ๊ฐ€์ค‘ํ‰๊ท  ์ ์‘ํ˜• ์ด์ง„ํ™” ์•Œ๊ณ ๋ฆฌ์ฆ˜์€ ๊ฐ ํ”ฝ์…€(Pixel) ์ฃผ๋ณ€์˜ (Block size) ร— (Block size) ์˜์—ญ์— ๋Œ€ํ•œ ๊ฐ€์ค‘ํ‰๊ท ์„ ๊ณ„์‚ฐํ•œ๋‹ค. ์ดํ›„ ๊ฐ€์ค‘ํ‰๊ท ์—์„œ ์ƒ์ˆ˜(Constant)๋ฅผ ๊ฐ์‚ฐํ•œ ๊ฐ’์„ ๊ณ„์‚ฐํ•ด์„œ ํ”ฝ์…€๋งˆ๋‹ค ์ ์‘ํ˜• ์ž„๊ณ„๊ฐ’์„ ์„ค์ •ํ•œ๋‹ค. ๊ฐ ํ”ฝ์…€์— ์ ์šฉ๋œ ์ ์‘ํ˜• ์ž„๊ณ„๊ฐ’์— ๋”ฐ๋ผ ์ž„๊ณ„๊ฐ’ ๋ณด๋‹ค ๋‚ฎ์€ ํ”ฝ์…€๊ฐ’์€ 0์œผ๋กœ ๋ณ€๊ฒฝํ•˜๊ณ , ์ž„๊ณ„๊ฐ’ ๋ณด๋‹ค ๋†’์€ ํ”ฝ์…€๊ฐ’์€ ์ตœ๋Œ“๊ฐ’์œผ๋กœ ๋ณ€๊ฒฝํ•œ๋‹ค. ๊ทธ๋ฆผ 7์€ ์ ์‘ํ˜• ์ด์ง„ํ™” ์•Œ๊ณ ๋ฆฌ์ฆ˜๋ฅผ ์ ์šฉํ•ด ๊ทธ๋ฆผ 6์˜ ๊ทธ๋ ˆ์ด์Šค์ผ€์ผ ์ด๋ฏธ์ง€๋ฅผ ๋ฐฑ์ƒ‰๊ณผ ๊ฒ€์ •์ƒ‰์˜ ์ด์ง„ ์ด๋ฏธ์ง€๋กœ ๋ณ€๊ฒฝํ•œ ๊ฒฐ๊ณผ์ด๋‹ค.

๊ทธ๋ฆผ. 6. ๊ฐ€์šฐ์‹œ์•ˆ ํ๋ฆผ ํšจ๊ณผ์™€ ํžˆํŠธ๋ฏธ์Šค ๋ชจํด๋กœ์ง€ ์—ฐ์‚ฐ ์ ์šฉ ์ „(์œ„)๊ณผ ํ›„(์•„๋ž˜)

Fig. 6. Before(up) and after(down) applying gaussian blur and hit-or-miss morphology operations

../../Resources/kiee/KIEE.2021.70.6.945/fig6.png

๊ทธ๋ฆผ. 7. ์ ์‘ํ˜• ์ด์ง„ํ™” ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์ ์šฉ ๊ฒฐ๊ณผ

Fig. 7. The result of applying binarization algorithms using adaptive threshold

../../Resources/kiee/KIEE.2021.70.6.945/fig7.png

3.4 ์›๊ทผ ์ ์šฉ

์นด๋ฉ”๋ผ๋ฅผ ํ†ตํ•ด ํš๋“ํ•œ ์ด๋ฏธ์ง€๋Š” ์ดฌ์˜ ๊ฐ๋„์— ๋”ฐ๋ผ ์ผ์ •ํ•œ ์›๊ทผ ์„ฑ๋ถ„์„ ํฌํ•จํ•˜๊ณ  ์žˆ๋‹ค. ์ด๋Š” ์˜ค๋ฅ˜ ์š”์†Œ๊ฐ€ ๋  ์ˆ˜ ์žˆ๋‹ค. ์›๊ทผ ์„ฑ๋ถ„์— ์˜ํ•œ ์˜ค๋ฅ˜ ์š”์†Œ๋ฅผ ์ œ๊ฑฐํ•˜๊ธฐ ์œ„ํ•ด ํš๋“๋œ ์ด๋ฏธ์ง€์— ํˆฌ์‹œ ๋ณ€ํ™˜(Perspective transform)์„ ์ ์šฉํ•œ๋‹ค. ํˆฌ์‹œ ๋ณ€ํ™˜์€ ์•„ํ™‰ ๊ฐœ์˜ ์›์†Œ๋ฅผ ๊ฐ–๋Š” 3ร—3 ํ–‰๋ ฌ์„ ์‚ฌ์šฉํ•œ๋‹ค. ํˆฌ์‹œ ๋ณ€ํ™˜์„ ํ‘œํ˜„ํ•˜๋Š” ํ–‰๋ ฌ์„ MP๋ผ๊ณ  ํ•˜๋ฉด, ์ž…๋ ฅ ์˜์ƒ์˜ ํ”ฝ์…€ ์ขŒํ‘œ (x,y)๊ฐ€ ํ–‰๋ ฌ MP์— ์˜ํ•ด ์ด๋™ํ•˜๋Š” ๊ฒฐ๊ณผ ์˜์ƒ ํ”ฝ์…€ ์ขŒํ‘œ (xโ€ฒ,yโ€ฒ)๋Š” ๋‹ค์Œ ์‹ (1)๊ณผ ๊ฐ™๋‹ค.

(1)
[ฮฑxโ€ฒฮฑyโ€ฒฮฑ]=M0[xy1]=[p11p12p13p21p22p23p31p32p33][xy1]

์‹ (1)์—์„œ ฮฑ๋Š” ๊ฒฐ๊ณผ ์˜์ƒ์˜ ์ขŒํ‘œ๋ฅผ ํ‘œํ˜„ํ•  ๋•Œ ์‚ฌ์šฉ๋˜๋Š” ๋น„๋ก€ ์ƒ์ˆ˜์ด๋ฉฐ ์‹ (2)์™€ ๊ฐ™๋‹ค. ๊ทธ๋Ÿฌ๋ฏ€๋กœ MP์— ์˜ํ•ด ์ด๋™ํ•˜๋Š” ๊ฒฐ๊ณผ ์˜์ƒ ํ”ฝ์…€ ์ขŒํ‘œ์˜ xโ€ฒ๊ณผ yโ€ฒ์€ ์‹ (3), ์‹ (4)์™€ ๊ฐ™๋‹ค.

(2)
ฮฑ=p31x+p32y+p33

(3)
xโ€ฒ=p11x+p12y+p13p31x+p32y+p33

(4)
yโ€ฒ=p21x+p22y+p23p31x+p32y+p33

์ด๋•Œ ํ–‰๋ ฌ MP๋Š” ์ž…๋ ฅ ์˜์ƒ์—์„œ ๋„ค ์ ์˜ ์ขŒํ‘œ์™€ ์ด ์ ๋“ค์ด ์ด๋™ํ•œ ๊ฒฐ๊ณผ ์˜์ƒ์˜ ์ขŒํ‘œ ๋„ค ๊ฐœ๋ฅผ ์ด์šฉํ•˜์—ฌ ๊ณ„์‚ฐํ•œ๋‹ค.

๊ทธ๋ฆผ 8์˜ ์œ„ ๊ทธ๋ฆผ์€ ์นด๋ฉ”๋ผ๋ฅผ ํ†ตํ•ด ํš๋“ํ•œ ์›๋ณธ ํƒ€์ผ ์ด๋ฏธ์ง€์ด๋ฉฐ, ๊ทธ๋ฆผ 8์˜ ์•„๋ž˜ ๊ทธ๋ฆผ์€ ์•„ํ•€(Affine) ์›๊ทผ ๋ณ€ํ™˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ ์šฉํ•ด ์›๋ณธ ์ด๋ฏธ์ง€๋กœ๋ถ€ํ„ฐ ์›๊ทผ ์š”์†Œ๋ฅผ ์ œ๊ฑฐํ•œ ์ด๋ฏธ์ง€์ด๋‹ค.

๊ทธ๋ฆผ. 8. ์•„ํ•€ ์›๊ทผ ๋ณ€ํ™˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์ ์šฉ ์ „(์œ„)๊ณผ ํ›„(์•„๋ž˜)

Fig. 8. Before(up) and after(down) applying affine and perspective transform algorithms

../../Resources/kiee/KIEE.2021.70.6.945/fig8.png

3.5 ๊ฐ€์žฅ์ž๋ฆฌ ๊ฒ€์ถœ๊ณผ ์ง์„  ๊ฒ€์ถœ

๊ทธ๋ฆผ 8์˜ ์•„๋ž˜ ์ด๋ฏธ์ง€๋กœ๋ถ€ํ„ฐ ์ด๋ฏธ์ง€ ๋‚ด์˜ ๊ฐ€์žฅ์ž๋ฆฌ ๊ฒ€์ถœ์„ ์œ„ํ•ด ์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜(Canny edge detection algorithms)์„ ์‚ฌ์šฉํ•œ๋‹ค. ์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์€ ํฌ๊ฒŒ ๋„ค ๊ฐœ์˜ ์—ฐ์‚ฐ ๊ณผ์ •์„ ํฌํ•จํ•œ๋‹ค. ๊ทธ๋ฆผ 9๋Š” ์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์— ํฌํ•จ๋œ 4๊ฐ€์ง€ ์—ฐ์‚ฐ ๊ณผ์ •์„ ๋ณด์—ฌ์ค€๋‹ค.

์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์€ ๋‹ค๋ฅธ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์— ๋น„ํ•ด ์—๋Ÿฌ์œจ์ด ๋‚ฎ๊ณ  ์—ฃ์ง€ ์ถ”์ถœ์ด ์ •ํ™•ํ•˜๊ธฐ ๋•Œ๋ฌธ์— ์ด๋ฏธ์ง€๋กœ๋ถ€ํ„ฐ ์—ฃ์ง€๋ฅผ ๊ฒ€์ถœํ•  ๋•Œ ๋งŽ์ด ์‚ฌ์šฉ๋˜๊ณ  ์žˆ๋‹ค(9). ๊ทธ๋ฆผ 10์€ ๊ทธ๋ฆผ 8์˜ ์•„๋ž˜ ์ด๋ฏธ์ง€(์•„ํ•€ ์›๊ทผ ๋ณ€ํ™˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์ ์šฉ ํ›„)์—์„œ ๋…ธ์ด์ฆˆ๋ฅผ ์ œ๊ฑฐํ•˜๊ณ  ์ด์ง„ํ™”ํ•œ ํ›„, ์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ ์šฉํ•œ ๊ฒฐ๊ณผ์ด๋‹ค.

๊ทธ๋ฆผ. 9. ์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์—ฐ์‚ฐ ๊ณผ์ •

Fig. 9. The operation procedure of Canny edge detection algorithms

../../Resources/kiee/KIEE.2021.70.6.945/fig9.png

๊ทธ๋ฆผ. 10. ์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์ ์šฉ ๊ฒฐ๊ณผ

Fig. 10. The result of applying Canny edge detection algorithms

../../Resources/kiee/KIEE.2021.70.6.945/fig10.png

์ผ€๋‹ˆ ์—ฃ์ง€ ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ˆ˜ํ–‰ํ•ด ์–ป์€ ๊ทธ๋ฆผ 10์˜ ์ด๋ฏธ์ง€๋กœ๋ถ€ํ„ฐ ํƒ€์ผ์˜ ์ง์„  ์„ฑ๋ถ„์„ ์ถ”์ถœํ•œ๋‹ค. ๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ํ‘œ์ค€ ํ—ˆํ”„ ๋ณ€ํ™˜(Standard Hough transform)์„ ์ด์šฉํ•œ ์ง์„  ์ถ”์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ด์šฉํ•œ๋‹ค.

ํ—ˆํ”„ ๋ณ€ํ™˜์€ 2์ฐจ์› xy์ขŒํ‘œ์—์„œ ์ง์„ ์˜ ๋ฐฉ์ •์‹์„ ํŒŒ๋ผ๋ฏธํ„ฐ ๊ณต๊ฐ„์œผ๋กœ ๋ณ€ํ™˜ํ•˜์—ฌ ์ง์„ ์„ ์ฐพ๋Š” ์•Œ๊ณ ๋ฆฌ์ฆ˜์ด๋‹ค(10). ์ผ๋ฐ˜์ ์œผ๋กœ ์ง์„ ์˜ ๋ฐฉ์ •์‹์€ ์‹ (5)์˜ ํ˜•ํƒœ๋กœ ์‚ฌ์šฉํ•œ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์ด์™€ ๊ฐ™์€ ์ง์„ ์˜ ๋ฐฉ์ •์‹์€ y์ถ•๊ณผ ํ‰ํ–‰ํ•œ ์ˆ˜์ง์„ ์„ ํ‘œํ˜„ํ•˜์ง€ ๋ชปํ•˜๊ธฐ ๋•Œ๋ฌธ์— ํ—ˆํ”„ ๋ณ€ํ™˜์„ ๊ตฌํ˜„ํ•  ๋•Œ๋Š” ์‹ (6)๊ณผ ๊ฐ™์€ ๊ทน์ขŒํ‘œ ํ˜•์‹์˜ ์ง์„ ์˜ ๋ฐฉ์ •์‹์„ ์‚ฌ์šฉํ•œ๋‹ค.

(5)
y=ax+b

(6)
xcosฮธ+ysinฮธ=ฯ

์œ„ ์‹ (6)์—์„œ ฯ๋Š” ์›์ ์—์„œ ์ง์„ ๊นŒ์ง€์˜ ์ˆ˜์ง ๊ฑฐ๋ฆฌ๋ฅผ ๋‚˜ํƒ€๋‚ด๊ณ , ฮธ๋Š” ์ง์„ ์— ์ˆ˜์ง์ด๋ฉฐ ์›์ ์„ ์ง€๋‚˜๋Š” ์„ ์ด x์ถ•๊ณผ ์ด๋ฃจ๋Š” ๊ฐ๋„๋ฅผ ์˜๋ฏธํ•œ๋‹ค. ์ด๋Š” xy๊ณต๊ฐ„์—์„œ ํ•œ ์ ์ด ฯฮธ๊ณต๊ฐ„์—์„œ ์‚ผ๊ฐํ•จ์ˆ˜ ๊ทธ๋ž˜ํ”„ ํ˜•ํƒœ์˜ ๊ณก์„ ์œผ๋กœ ํ‘œํ˜„๋˜๊ณ , ฯฮธ๊ณต๊ฐ„์—์„œ์˜ ํ•œ ์ ์€ xy๊ณต๊ฐ„์—์„œ ์ง์„ ์˜ ํ˜•ํƒœ๋กœ ๋‚˜ํƒ€๋‚ผ ์ˆ˜ ์žˆ์Œ์„ ๋ณด์—ฌ์ค€๋‹ค. ๊ทธ๋ฆผ 11์€ ฯฮธํŒŒ๋ผ๋ฏธํ„ฐ๋ฅผ ์ด์šฉํ•œ ์ง์„  ํ‘œํ˜„์‹ ์˜ˆ์‹œ ๊ทธ๋ž˜ํ”„์ด๋‹ค. ๊ทธ๋ฆผ 11์˜ (a)๋Š” 2์ฐจ์› xy ์ขŒํ‘œ๊ณ„์ด๋ฉฐ, ์ง์„ ์˜ ๋ฐฉ์ •์‹์€ xcosฮธ+ysinฮธ=ฯ์ด๋‹ค. ์ด ์ง์„  ์œ„์˜ ๋‘ ์ ์— ๋Œ€ํ•˜์—ฌ ๊ฐ ์ ์— ๋Œ€์‘ํ•˜๋Š” ฯฮธ๊ณต๊ฐ„์—์„œ์˜ ๊ณก์„ ์€ ๊ทธ๋ฆผ 11์˜ (b)์™€ ๊ฐ™๋‹ค. ฯฮธ๊ณต๊ฐ„์—์„œ ๋‘ ๊ณก์„ ์€ ํ•˜๋‚˜์˜ ์ ์—์„œ ๊ต์ฐจํ•˜๋ฉฐ, ์ด ์ ์˜ ์ขŒํ‘œ (ฯ0,ฮธ0)๊ฐ€ ๊ทธ๋ฆผ 11(a)์˜ ์ง์„ ์„ ๋‚˜ํƒ€๋‚ด๋Š” ํŒŒ๋ผ๋ฏธํ„ฐ๊ฐ€ ๋œ๋‹ค.

๊ทธ๋ฆผ. 11. ฯฮธํŒŒ๋ผ๋ฏธํ„ฐ๋ฅผ ์ด์šฉํ•œ ์ง์„  ํ‘œํ˜„์‹ ์˜ˆ์‹œ ๊ทธ๋ž˜ํ”„

Fig. 11. Example graph of linear expression using ฯฮธ parameters

../../Resources/kiee/KIEE.2021.70.6.945/fig11.png

ํ—ˆํ”„ ๋ณ€ํ™˜์„ ์ด์šฉํ•˜์—ฌ ์ง์„ ์„ ์ฐพ์œผ๋ ค๋ฉด xy ๊ณต๊ฐ„์—์„œ ์—ฃ์ง€๋กœ ํŒ๋ณ„๋œ ๋ชจ๋“  ์ ์„ ์ด์šฉํ•˜์—ฌ ฯฮธ ํŒŒ๋ผ๋ฏธํ„ฐ ๊ณต๊ฐ„์— ๊ณก์„ ์œผ๋กœ ํ‘œํ˜„ํ•˜๊ณ , ์ถ•์  ๋ฐฐ์—ด์„ ์ด์šฉํ•˜์—ฌ ๊ณก์„ ์ด ๋งŽ์ด ๊ต์ฐจ๋˜๋Š” ์ ์„ ์ฐพ๋Š”๋‹ค. ์ด ์ ์€ xy ๊ณต๊ฐ„์—์„œ์˜ ์„ ์„ ์˜๋ฏธํ•œ๋‹ค.

๊ทธ๋ฆผ 12๋Š” ์—ฃ์ง€ ๊ฒ€์ถœ๊ณผ ํ—ˆํ”„ ๋ณ€ํ™˜ ์ง์„  ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ํ†ตํ•ด ๊ฒ€์ถœ๋œ ์ง์„ ์„ ์›๋ณธ ์ด๋ฏธ์ง€ ์œ„์— ๋„์‹œํ•œ ๊ทธ๋ฆผ์ด๋‹ค.

๊ทธ๋ฆผ. 12. ํ—ˆํ”„ ๋ณ€ํ™˜ ์ง์„  ๊ฒ€์ถœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ํ†ตํ•ด ๊ฒ€์ถœ๋œ ์ง์„ 

Fig. 12. Straight lines detected through line detection algorithms using hough transform

../../Resources/kiee/KIEE.2021.70.6.945/fig12.png

๊ทธ๋ฆผ. 13. ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์ œ์–ด๋ฅผ ์œ„ํ•œ ์ˆ˜ํ‰ ๋ฐ ์ˆ˜์ง ๊ฐ€์ด๋“œ๋ผ์ธ

Fig. 13. Horizontal and vertical guideline for driving control of autonomous mobile robot tracks tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig13.png

3.6 ๊ฐ€์ด๋“œ๋ผ์ธ ์ถ”์ถœ

๊ทธ๋ฆผ 12์˜ ์ด๋ฏธ์ง€์— ์žˆ๋Š” ๋‹ค์ˆ˜์˜ ์„ ๋“ค ์ค‘์—์„œ ์ด๋ฏธ์ง€์˜ ์ค‘์‹ฌ์„ 2์ฐจ์› ํ‰๋ฉด์˜ ์›์ ์œผ๋กœ ๊ฐ€์ •ํ•  ๋•Œ, ์ง์„ ์˜ ๊ธฐ์šธ๊ธฐ๋ฅผ a๋ผ ํ•˜๋ฉด, a์˜ ํฌ๊ธฐ๊ฐ€ โˆ’1โ‰คaโ‰ค1์ธ ์„ ๊ณผ ๊ทธ ์™ธ์˜ ์„ ๋“ค๋กœ ๋ถ„๋ฆฌํ•  ์ˆ˜ ์žˆ๋‹ค. ๊ธฐ์šธ๊ธฐ์— ๋”ฐ๋ผ ๋‘ ์ข…๋ฅ˜๋กœ ๋ถ„๋ฅ˜๋œ ๊ฐ๊ฐ์˜ ์„ ๋“ค์— ๋Œ€ํ•˜์—ฌ ์„ ์˜ ๊ธธ์ด์™€ ๊ธฐ์šธ๊ธฐ์˜ ์œ ์‚ฌ๋„๋ฅผ ๋น„๊ตํ•˜์—ฌ ๊ธฐ์šธ๊ธฐ๊ฐ€ ์œ ์‚ฌํ•œ ์„  ์ค‘ ๊ฐ€์žฅ ๊ธธ์ด๊ฐ€ ๊ธด ๋‘ ์ข…๋ฅ˜์˜ ์„ ์„ ์ถ”์ถœํ•œ๋‹ค. ํƒ€์ผ์˜ ํฌ๊ธฐ๊ฐ€ ์ž‘์•„ ์ด๋ฏธ์ง€์— ๋‹ค์ˆ˜์˜ ํƒ€์ผ ์žˆ๋Š” ๊ฒฝ์šฐ์—๋Š” ์›์ ์—์„œ ๊ฐ€์žฅ ๊ฐ€๊นŒ์šด ํƒ€์ผ์˜ ๊ฒฉ์ž๋ฌด๋Šฌ๋ฅผ ๊ธฐ์ค€์œผ๋กœ ์ง์„  ์ถ”์ถœ ๊ณผ์ •์„ ์ˆ˜ํ–‰ํ•œ๋‹ค. ๊ทธ๋ฆผ 13์€ ๊ทธ๋ฆผ 12์˜ ๋‹ค์ˆ˜์˜ ์„ ๋“ค ์ค‘์—์„œ ์„ ์˜ ๊ธฐ์šธ๊ธฐ๋ฅผ ๋Œ€ํ‘œํ•˜๊ณ  ๊ฐ€์žฅ ๊ธธ์ด๊ฐ€ ๊ธด ์„ ๋งŒ ์ถ”์ถœํ•œ ํ›„ ์›๋ณธ ์ด๋ฏธ์ง€์˜ ๊ฐ€์žฅ์ž๋ฆฌ๊นŒ์ง€ ์„ ์„ ์—ฐ์žฅํ•˜์—ฌ ๊ทธ๋ฆฐ ์ด๋ฏธ์ง€์ด๋‹ค. ์ด ๋‘ ๊ฐœ์˜ ์„ ์€ ์ด๋™ ๋กœ๋ด‡์˜ ํ˜„์žฌ์˜ ์œ„์น˜์™€ ์ฃผํ–‰ ๋ฐฉํ–ฅ์„ ์˜๋ฏธํ•˜๋ฉฐ, ์ฃผํ–‰ ์ œ์–ด๋ฅผ ์œ„ํ•œ ์ˆ˜์ง, ์ˆ˜ํ‰ ๊ฐ€์ด๋“œ๋ผ์ธ์œผ๋กœ ์‚ฌ์šฉ๋œ๋‹ค. ํŠนํžˆ ์ˆ˜ํ‰ ๊ฐ€์ด๋“œ๋ผ์ธ์€ ์ฃผํ–‰ ๊ฑฐ๋ฆฌ๋ฅผ ์ธก์ •ํ•˜๊ธฐ ์œ„ํ•œ ์—”์ฝ”๋”(Encoder)๋กœ ํ™œ์šฉ๋œ๋‹ค.

4. ์ด๋™ ๋กœ๋ด‡ ์ฃผํ–‰ ์ œ์–ด

4.1 ์ฃผํ–‰ ์ œ์–ด ์‹œ์Šคํ…œ ์„ค๊ณ„

๋ณธ ๋…ผ๋ฌธ์—์„œ ์ด๋™ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์ œ์–ด๋ฅผ ์œ„ํ•ด ์‚ฌ์šฉ๋  ์ฃผํ–‰ ์ œ์–ด๊ธฐ์˜ ๊ตฌ์กฐ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค. ๊ทธ๋ฆผ 14๋Š” ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์ œ์–ด๋ฅผ ์œ„ํ•œ PID ์ œ์–ด ์‹œ์Šคํ…œ์˜ ๋ธ”๋ก๋„์ด๋‹ค.

์กฐ์ž‘๋Ÿ‰ u(t)๋Š” ์—ฐ์†์‹ ํ˜ธ๋กœ์„œ ์•„๋ž˜ ์‹ (7)๊ณผ ๊ฐ™๋‹ค. u(t)๋Š” ๋ชฉํ‘œ๊ฐ’๊ณผ ์ธก์ •๊ฐ’ ์‚ฌ์ด์˜ ์—๋Ÿฌ e(t)์— ๋Œ€ํ•œ ๋น„๋ก€(Proportional)ํ•ญ. ์ ๋ถ„(Integral)ํ•ญ, ๋ฏธ๋ถ„(Differential)ํ•ญ์˜ ํ•ฉ์œผ๋กœ ์ด๋ฃจ์–ด์ง„๋‹ค. ๊ฐ ํ•ญ์—๋Š” ๋น„๋ก€์ƒ์ˆ˜(Kp), ์ ๋ถ„์ƒ์ˆ˜(Ki), ๋ฏธ๋ถ„์ƒ์ˆ˜(Kd)๋ฅผ ์‚ฌ์šฉํ•œ๋‹ค.

๊ทธ๋ฆผ. 14. ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์ œ์–ด๋ฅผ ์œ„ํ•œ PID ์ œ์–ด ์‹œ์Šคํ…œ ๋ธ”๋ก๋„

Fig. 14. Block diagram of PID control system for driving control of autonomous mobile robot that tracks tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig14.png

(7)
u(t)=Kpe(t)+Kiโˆซt0e(ฯ„)dฯ„+Kdde(t)dt

์‹ (7)์„ ๋ถˆ์—ฐ์† ์‹ ํ˜ธ๋กœ ๋ณ€๊ฒฝํ•˜๋ฉด ์‹ (8)์™€ ๊ฐ™๋‹ค. ์‹ (8)์˜ ์กฐ์ž‘๋Ÿ‰ u[n]์€ n๋ฒˆ์งธ ์ธก์ •๊ฐ’์œผ๋กœ, ๊ณ„์‚ฐ๋œ ๋ชฉํ‘œ๊ฐ’๊ณผ ์ธก์ •๊ฐ’ ์‚ฌ์ด์˜ ์—๋Ÿฌ e[n]์„ ์‚ฌ์šฉํ•˜์—ฌ ๋งŒ๋“  ์ œ์–ด ์‹ ํ˜ธ๋‹ค. โ–ณt๋Š” ์—๋Ÿฌ๊ฐ’์„ ์ธก์ •ํ•˜๊ณ  ์กฐ์ž‘๋Ÿ‰์„ ์ธ๊ฐ€ํ•˜๋Š” ์ฃผ๊ธฐ๋ฅผ ๋‚˜ํƒ€๋‚ธ๋‹ค. ๋ณธ ๋…ผ๋ฌธ์—์„œ ๊ฐํ•ญ์˜ ๊ณ„์ˆ˜ Kp, Ki, Kd๋Š” ๊ฒฝํ—˜์น˜๋กœ ๊ตฌํ•˜์˜€๋‹ค.

(8)
u[n]=Kpe[n]+Kiโˆ‘ni=0e[n]โ–ณt+Kde[n]โˆ’e[nโˆ’1]โ–ณt

4.2 ์ฃผํ–‰ ์ œ์–ด

๊ทธ๋ฆผ 15๋Š” ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ๋ฐฉํ–ฅ ์ œ์–ด๋ฅผ ์œ„ํ•œ ๋ฐฉ๋ฒ•์„ ์„ค๋ช…ํ•˜๊ธฐ ์œ„ํ•œ ๊ทธ๋ž˜ํ”„์ด๋‹ค. ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์‹œ์ž‘์ ์—์„œ ํš๋“ํ•œ ์ˆ˜ํ‰ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์ˆ˜์ง ๊ฐ€์ด๋“œ๋ผ์ธ์„ ์ดˆ๊ธฐ ๊ฐ€์ด๋“œ๋ผ์ธ(Initial Guideline)์ด๋ผ ํ•˜๊ณ , ๋กœ๋ด‡์ด ์ฃผํ–‰์„ ์‹œ์ž‘ํ•œ ํ›„์— ํš๋“ํ•œ ์ˆ˜ํ‰ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์ˆ˜์ง ๊ฐ€์ด๋“œ๋ผ์ธ์„ ์ฃผํ–‰์‹œ๊ฐ„ ๊ฐ€์ด๋“œ๋ผ์ธ(Runtime Guideline)์ด๋ผ ์ •ํ•œ๋‹ค. ๊ทธ๋ž˜ํ”„์—์„œ O๋Š” ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์‹œ์ž‘์ ์—์„œ ํš๋“ํ•œ ํƒ€์ผ ์˜์ƒ์˜ ์ˆ˜ํ‰ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์ˆ˜์ง ๊ฐ€์ด๋“œ๋ผ์ธ์˜ ๊ต์ฐจ์ ์ด๋‹ค. ์ด ๊ต์ฐจ์ ์„ ์›์ ์ด๋ผ ํ•œ๋‹ค. ์ฃผํ–‰ ์‹œ์ž‘ ํ›„ ํš๋“ํ•œ ์ด๋ฏธ์ง€๋กœ๋ถ€ํ„ฐ ์ถ”์ถœ๋œ ์ˆ˜ํ‰ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์ˆ˜์ง ๊ฐ€์ด๋“œ๋ผ์ธ์ด ์›์ ์„ ์ง€๋‚˜๋Š” ์ˆ˜ํ‰์„ ๊ณผ ๋งŒ๋‚˜๋Š” ์ ์„ Xh์™€ Xv๋ผ๊ณ  ํ•œ๋‹ค.

์ž„์˜์˜ ์‹œ๊ฐ„์— ์ˆ˜ํ‰ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์›์ ์„ ์ง€๋‚˜๋Š” ์ˆ˜ํ‰์„ ์ด ์ด๋ฃจ๋Š” ๊ฐ์„ ฮธ(t), ์ˆ˜์ง ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์›์ ์„ ์ง€๋‚˜๋Š” ์ˆ˜ํ‰์„ ์˜ ๊ต์  Xv์™€ ์›์ ๊นŒ์ง€์˜ ๊ฑฐ๋ฆฌ๋ฅผ d(t)๋ผ ํ•˜๋ฉด, ๋กœ๋ด‡์˜ ์ดˆ๊ธฐ ์œ„์น˜์™€ ํ˜„ ์œ„์น˜ ์‚ฌ์ด์˜ ์ฐจ์ด e(t)๋ฅผ ์‹ (9)๊ณผ ๊ฐ™์ด ์ •์˜ํ•  ์ˆ˜ ์žˆ๋‹ค. ์‹ (9)์—์„œ ๊ฐ€์ค‘๊ฐ’์€ ๊ฒฝํ—˜์น˜๋กœ ๊ตฌํ•  ์ˆ˜ ์žˆ๋‹ค.

๊ทธ๋ฆผ. 15. ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์ œ์–ด๋ฅผ ์œ„ํ•œ ํŒŒ๋ผ๋ฏธํ„ฐ

Fig. 15. Parameters for driving control of autonomous mobile robot tracking tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig15.png

(9)
e(t)=ฮธ(t)+Wd(t), W:๊ฐ€์ค‘๊ฐ’(WeightedValue)

์ด๋ฅผ ์ด๋ฏธ์ง€์˜ ํŠน์ • ํ”„๋ ˆ์ž„์— ๋Œ€ํ•ด ๊ณ„์‚ฐํ•˜๊ธฐ ์œ„ํ•ด ์ด์‚ฐ ์‹์œผ๋กœ ๋ฐ”๊พธ๋ฉด ์‹ (10)์ด ๋˜๋ฉฐ, ์ฃผํ–‰ ๋กœ๋ด‡์ด ๋ฐ”๋‹ฅ ํƒ€์ผ์„ ๋”ฐ๋ผ ์ง์„  ์ฃผํ–‰์„ ํ•˜๊ธฐ ์œ„ํ•œ ์กฐ๊ฑด์‹์€ ์‹ (11)์™€ ๊ฐ™๋‹ค.

(10)
e[n]=ฮธ[n]+Wd[n], W:๊ฐ€์ค‘๊ฐ’(WeightedValue)

(11)
[ if e[n]=0 then RPM_R_Wheel = RPM_L_Wheel  if e[n]>0 then RPM_R_Wheel > RPM_L_Wheel  if e[n]<0 then RPM_R_Wheel < RPM_L_Wheel  where RPM_R_Wheel is RPM of right wheel,  RPM_L_Wheel is RPM of left wheel ]

์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ดˆ๊ธฐ ๊ตฌ๋™ ์‹œ ์ขŒ์šฐ ๋ชจํ„ฐ๋ฅผ ๊ตฌ๋™ํ•˜๊ธฐ ์œ„ํ•œ PWM(Pulse Width Modulation) ๋“€ํ‹ฐ๋น„(Duty ratio)๋Š” ๋™์ผํ•˜๋‹ค. ์ฃผํ–‰ ๋กœ๋ด‡์ด ์ง์„  ์ฃผํ–‰์„ ํ•˜๊ธฐ ์œ„ํ•ด์„œ๋Š” e[n]์ด 0๋ณด๋‹ค ํฌ๋ฉด ์ขŒ์ธก ๋ชจํ„ฐ์˜ RPM์„ ๊ณ ์ •ํ•˜๊ณ  PID ์ œ์–ด ์•Œ๊ณ ๋ฆฌ์ฆ˜์— ๋”ฐ๋ผ ์šฐ์ธก ๋ชจํ„ฐ๋ฅผ ์ œ์–ดํ•œ๋‹ค. e[n]์ด 0๋ณด๋‹ค ์ž‘์œผ๋ฉด ๊ทธ ๋ฐ˜๋Œ€ ๋ฐฉ๋ฒ•์œผ๋กœ ์ œ์–ด๋ฅผ ์ˆ˜ํ–‰ํ•œ๋‹ค.

5. ์‹คํ—˜ ๊ฒฐ๊ณผ

๋ณธ ๋…ผ๋ฌธ์—์„œ ๊ตฌํ˜„ํ•œ ์‹ค๋‚ด ์ž์œจ ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์„ฑ๋Šฅ์„ ํ‰๊ฐ€ํ•˜๊ธฐ ์œ„ํ•œ ์‹คํ—˜์„ ์ˆ˜ํ–‰ํ•˜์˜€๋‹ค. ์‹คํ—˜ ํ™˜๊ฒฝ์€ ๋ฐ”๋‹ฅ์— ํƒ€์ผ์ด ์„ค์น˜๋œ ์‹ค๋‚ด ๊ณต๊ฐ„์ด๋ฉฐ, ์‹คํ—˜ ๊ณต๊ฐ„์˜ ์ฒœ์žฅ๊ณผ ๋ฒฝ๋ฉด์—๋Š” ๋‹ค์ˆ˜์˜ ์กฐ๋ช…๊ณผ ํˆฌ๋ช…์ฐฝ์ด ์„ค์น˜๋˜์–ด ์žˆ๋‹ค.

๊ทธ๋ฆผ 16์€ ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์„ฑ๋Šฅ ํ‰๊ฐ€๋ฅผ ์œ„ํ•œ ์‹คํ—˜ ๋ฐฉ๋ฒ•์„ ๋ณด์—ฌ์ฃผ๋Š” ๊ทธ๋ฆผ์ด๋‹ค. ๋กœ๋ด‡์€ ์ •ํ•ด์ง„ ์‹œ์ž‘์ (Start)์—์„œ ์ถœ๋ฐœํ•˜์—ฌ ์ •ํ•ด์ง„ ์ •์ง€ ์œ„์น˜(End)๊นŒ์ง€ ์ง์„  ์ฃผํ–‰ ํ›„, ์—ญ์ฃผํ–‰ํ•˜์—ฌ ๋‹ค์‹œ ์‹œ์ž‘์ ๊นŒ์ง€ ๋Œ์•„์˜ค๋„๋ก ํ”„๋กœ๊ทธ๋ž˜๋ฐ๋˜์—ˆ๋‹ค. ์ด๋•Œ ์ฃผํ–‰ ๊ฑฐ๋ฆฌ๋Š” ๋กœ๋ด‡์ด ์ธ์‹ํ•œ ํƒ€์ผ์˜ ๊ฐœ์ˆ˜๋ฅผ ์—”์ฝ”๋”ฉํ•˜์—ฌ ๊ฑฐ๋ฆฌ๋กœ ํ™˜์‚ฐํ•˜๋„๋ก ํ•˜์˜€๋‹ค.

๊ทธ๋ฆผ. 16. ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์„ฑ๋Šฅ ํ‰๊ฐ€๋ฅผ ์œ„ํ•œ ์‹คํ—˜ ๋ฐฉ๋ฒ•

Fig. 16. Experimental method for evaluating driving performance of autonomous mobile robot tracking tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig16.png

๋‹ค์Œ ๊ทธ๋ฆผ 17์€ ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์„ฑ๋Šฅ ํ‰๊ฐ€ ์‹คํ—˜ ๊ฒฐ๊ณผ์ด๋‹ค. ์‹คํ—˜ ๊ฒฐ๊ณผ ๊ทธ๋ž˜ํ”„์—์„œ๋Š” ์‹œ์ž‘์ ์—์„œ ์ถœ๋ฐœํ•˜์—ฌ ์ •ํ•ด์ง„ ๋ชฉํ‘œ์ ์— ๋„์ฐฉ ํ›„ ์—ญ์ฃผํ–‰ํ•˜์—ฌ ๋‹ค์‹œ ์‹œ์ž‘์ ์— ๋„์ฐฉํ–ˆ์„ ๋•Œ ๋กœ๋ด‡์˜ ์ดˆ๊ธฐ ์œ„์น˜ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ํ˜„ ์œ„์น˜ ๊ฐ€์ด๋“œ๋ผ์ธ ์‚ฌ์ด์˜ ์˜ค์ฐจ ๊ฐ ฮธ์™€ ์ˆ˜์ง ๊ฐ€์ด๋“œ๋ผ์ธ์˜ ์˜ค์ฐจ ๊ฑฐ๋ฆฌ d๋ฅผ ๋ณด์—ฌ์ค€๋‹ค.

๋ณธ ๋…ผ๋ฌธ์—์„œ ๊ตฌํ˜„ํ•œ ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์„ฑ๋Šฅ ์‹คํ—˜์„ ์ˆ˜ํ–‰ํ•œ ๊ฒฐ๊ณผ ๊ทธ๋ž˜ํ”„์—์„œ D1๊ณผ ฮธ1์€ 12m ์™•๋ณต ์ฃผํ–‰ ์‹คํ—˜์—์„œ ์ดˆ๊ธฐ ์œ„์น˜ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์ฃผํ–‰ ์ข…๋ฃŒ ์‹œ์  ํ˜„ ์œ„์น˜ ๊ฐ€์ด๋“œ๋ผ์ธ ์‚ฌ์ด์˜ ์˜ค์ฐจ ๊ฐ(ฮธ)๊ณผ ์˜ค์ฐจ ๊ฑฐ๋ฆฌ(d)์ด๋ฉฐ, D2๊ณผ ฮธ2์€ 24m ์™•๋ณต ์ฃผํ–‰ ์‹คํ—˜์—์„œ ์ดˆ๊ธฐ ์œ„์น˜ ๊ฐ€์ด๋“œ๋ผ์ธ๊ณผ ์ฃผํ–‰ ์ข…๋ฃŒ ์‹œ์  ํ˜„ ์œ„์น˜ ์ตœ์ข… ๊ฐ€์ด๋“œ๋ผ์ธ ์‚ฌ์ด์˜ ์˜ค์ฐจ ๊ฐ(ฮธ)๊ณผ ์˜ค์ฐจ ๊ฑฐ๋ฆฌ(d)์ด๋‹ค. ๊ฐ ๊ฑฐ๋ฆฌ์— ๋Œ€ํ•œ ์ด 20ํšŒ์˜ ์™•๋ณต ์ฃผํ–‰ ๊ฒฐ๊ณผ์—์„œ 12m ์™•๋ณต ์ฃผํ–‰ ์‹œ ์ตœ๋Œ€ ์˜ค์ฐจ ๊ฐ(ฮธmax12)๊ณผ ์˜ค์ฐจ ๊ฑฐ๋ฆฌ(dmax12)๋Š” 9.3ยฐ์™€ 87mm์ด๊ณ , 24m ์™•๋ณต ์ฃผํ–‰ ์‹œ ์ตœ๋Œ€ ์˜ค์ฐจ ๊ฐ(ฮธmax24)๊ณผ ์˜ค์ฐจ ๊ฑฐ๋ฆฌ(dmax24)๋Š” -9.6ยฐ์™€ 88mm์ด๋‹ค. ์ด์™€ ๊ฐ™์€ ๊ฒฐ๊ณผ๋Š” ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ง์„  ์ฃผํ–‰ ์„ฑ๋Šฅ์ด ๋งค์šฐ ์šฐ์ˆ˜ํ•จ์„ ๋ณด์—ฌ์ค€๋‹ค.

๊ทธ๋ฆผ. 17. ํƒ€์ผ ๊ฒฉ์ž ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์ฃผํ–‰ ์„ฑ๋Šฅ ํ‰๊ฐ€ ์‹คํ—˜ ๊ฒฐ๊ณผ

Fig. 17. Experiment results for evaluating driving performance of autonomous mobile robot tracking tile grid pattern

../../Resources/kiee/KIEE.2021.70.6.945/fig17.png

์‹คํ—˜์ด ์‹ค์‹œ๋œ ์‹ค๋‚ด ๋ฐ”๋‹ฅ์— ์„ค์น˜๋œ ํƒ€์ผ์˜ ํฌ๊ธฐ๋Š” 60cmร—60cm์ด๋‹ค. ๊ทธ๋Ÿฌ๋ฏ€๋กœ ์‹œ์ž‘ ์œ„์น˜์™€ ์ •์ง€ ์œ„์น˜ ์‚ฌ์ด์— ์„ค์น˜๋œ ํƒ€์ผ์˜ ๊ฐœ์ˆ˜๋Š” ์ฃผํ–‰ ๊ฑฐ๋ฆฌ 12m๋Š” 20๊ฐœ, ์ฃผํ–‰ ๊ฑฐ๋ฆฌ 24m๋Š” 40๊ฐœ์ด๋‹ค. ์‹คํ—˜ ๊ฒฐ๊ณผ ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์€ ์ฃผํ–‰ ์ค‘ ์ธ์‹๋œ ํƒ€์ผ์˜ ์ˆ˜๋ฅผ ์ •ํ™•ํ•˜๊ฒŒ ์นด์šดํŠธํ•˜๋Š” ๊ฒƒ์„ ํ™•์ธํ•˜์˜€๋‹ค.

๋ณธ ๋…ผ๋ฌธ์—์„œ ๊ตฌํ˜„ํ•œ ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์‹ค๋‚ด ์ž์œจ ์ฃผํ–‰ ๋กœ๋ด‡์€ ์ฃผํ–‰ ์ค‘ ๋ฐœ์ƒํ•˜๋Š” ๋ˆ„์  ์˜ค์ฐจ๊ฐ€ ํƒ€์ผ์ด ๋ฐ”๋€” ๋•Œ๋งˆ๋‹ค ์†Œ๊ฑฐ๋˜์–ด ์›๊ฑฐ๋ฆฌ, ์žฅ์‹œ๊ฐ„ ์ฃผํ–‰์—๋„ ์ฃผํ–‰ ๊ฒฝ๋กœ๋ฅผ ์ดํƒˆํ•˜์ง€ ์•Š๊ณ  ์ •ํ™•ํ•œ ๊ฒฝ๋กœ๋ฅผ ์ฃผํ–‰ํ•˜๋Š” ๊ฒƒ์„ ์‹คํ—˜ ๊ฒฐ๊ณผ๋ฅผ ํ†ตํ•ด ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค.

6. ๊ฒฐ ๋ก 

์ฃผํ–‰ ๋กœ๋ด‡์˜ ์œ„์น˜๋ฅผ ์ธ์‹์„ ํ•˜๋ ค๋Š” ์‹œ๋„๋Š” ์•„์ฃผ ์˜ค๋žœ ๊ธฐ๊ฐ„ ์ง„ํ–‰๋˜์–ด ์™”๋‹ค. ํ•˜์ง€๋งŒ ๋กœ๋ด‡์˜ ์ •ํ™•ํ•œ ์œ„์น˜๋ฅผ ์ธ์‹ํ•˜๋Š” ๊ธฐ์ˆ ์—๋Š” ๋งŽ์€ ์–ด๋ ค์›€์ด ์žˆ๋‹ค. ํ˜„์žฌ ๋กœ๋ด‡์˜ ์œ„์น˜ ์ธ์‹๊ณผ ๋„ค๋น„๊ฒŒ์ด์…˜(Navigation) ๊ธฐ์ˆ ์€ ์ธ๊ณต์ง€ํ‘œ๋‚˜ ๋น„์ „(Vision)์„ ๊ธฐ๋ฐ˜์œผ๋กœ ์ง„ํ–‰๋˜๊ณ  ์žˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด์™€ ๊ฐ™์€ ๊ธฐ์ˆ ์€ ์œ„์น˜ ์ธ์‹์˜ ์˜ค์ฐจ ๋ฒ”์œ„๊ฐ€ ์ผ์ • ๊ธฐ์ค€ ์ด์ƒ ๋ฐœ์ƒํ•˜๊ณ  ๊ธฐ์ดˆ ์‹œ๊ณต๋น„์šฉ์ด ๋†’์•„ ์ •ํ™•ํ•œ ์œ„์น˜ ์ธ์‹๊ณผ ๋‚ฎ์€ ์‹œ๊ณต๋น„์šฉ์ด ์š”๊ตฌ๋˜๋Š” ๊ธฐ์ˆ ์„ ํ•„์š”๋กœ ํ•œ๋‹ค(1).

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋กœ๋ด‡์ด ์ธ๊ฐ„์˜ ์‹œ๊ฐ ์ง€๋Šฅ์„ ๋ชจ๋ฐฉํ•˜์—ฌ ์‹ค๋‚ด ๊ณต๊ฐ„์˜ ๋ฐ”๋‹ฅ ๋ฉด์— ์„ค์น˜๋œ ํƒ€์ผ์˜ ๊ฒฉ์ž๋ฌด๋Šฌ ์ •๋ณด๋ฅผ ๋ถ„์„ํ•ด์„œ ์œ„์น˜๋ฅผ ์ธ์‹ํ•˜๊ณ  ์ฃผํ–‰ ๋ฐฉํ–ฅ์„ ๊ฒฐ์ •ํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์ œ์•ˆํ•˜์˜€๋‹ค. ์ œ์•ˆ๋œ ๋ฐฉ๋ฒ•์œผ๋กœ ์ œ์ž‘๋œ ์‹ค๋‚ด์ž์œจ ์ฃผํ–‰ ๋กœ๋ด‡์˜ ์„ฑ๋Šฅ ํ‰๊ฐ€๋ฅผ ์œ„ํ•ด ์ˆ˜ํ–‰ํ•œ ์ฃผํ–‰ ์‹คํ—˜ ๊ฒฐ๊ณผ์— ์˜ํ•ด ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋กœ๋ด‡์€ ๋ฐ”๋‹ฅ์— ํƒ€์ผ์ด ์„ค์น˜๋œ ์‹ค๋‚ด ๊ณต๊ฐ„์—์„œ ๋กœ๋ด‡์— ๋ถ€์ฐฉ๋œ ๋‹จ์ผ ์นด๋ฉ”๋ผ๋งŒ์œผ๋กœ ์•ฝ์†๋œ ์ฃผํ–‰ ๊ฒฝ๋กœ๋ฅผ ๋ˆ„์  ์˜ค์ฐจ ์—†์ด ์žฅ์‹œ๊ฐ„ ์ฃผํ–‰ํ•  ์ˆ˜ ์žˆ์Œ์„ ํ™•์ธํ•˜์˜€๋‹ค. ๋ณธ ๋…ผ๋ฌธ์—์„œ ์ œ์•ˆ๋œ ํƒ€์ผ ๊ฒฉ์ž๋ฌด๋Šฌ ์ถ”์  ์ž์œจ์ฃผํ–‰ ๋ฐฉ๋ฒ•์€ ๋ณ„๋„์˜ ์‹œ๊ณต๋น„์šฉ์ด ์—†์œผ๋ฉฐ ์ •ํ™•ํ•œ ์œ„์น˜ ์ธ์‹๊ณผ ๋ˆ„์  ์˜ค์ฐจ๊ฐ€ ์—†๋Š” ์ฃผํ–‰ ์„ฑ๋Šฅ์„ ๊ฐ–๊ณ  ์žˆ์œผ๋ฏ€๋กœ ํ–ฅํ›„ ๋‹ค์–‘ํ•œ ๋ถ„์•ผ์—์„œ ํ™œ์šฉ์ด ๊ฐ€๋Šฅํ•  ๊ฒƒ์œผ๋กœ ๊ธฐ๋Œ€๋œ๋‹ค.

References

1 
Su-Yong An, Jeong-Gwan Kang, Lae-Kyeong Lee, 2010, SLAM with Visually Salient Line Feature in Indoor Hallway Environments, Journal of Institute of Control, Robotics and Systems, Vol. 16, No. 1, pp. 40-47DOI
2 
Se-Jun Park, 2013, A Study on Design of Intelligent Mobile Robot based on Localization Sensor for Unmanned Transport Logistics, The Journal of Korean Institute of Information Technology, Vol. 11, No. 9, pp. 7-13Google Search
3 
Ki-Sung You, 2011, Development of Localization Sensor System for Intelligent Robot, Journal of Institute of Control Robotics and Systems, Vol. 17, No. 2, pp. 116-124DOI
4 
Andrea. Garuli, Antonio. Giannitrapani, Andrea. Rossi, Antonio. Vicino, 2005, Mobile robot SLAM for line-based environment representation, IEEE Conference on Decision and Control and the European Control Conference, pp. 2041-2046DOI
5 
V. J. Lumelsky, T. Skewis, 1990, Incorporating range sensing in the robot navigation function, IEEE Transactions on Systems Man and Cybernetics, Vol. 20, No. 5, pp. 1058-1069DOI
6 
Young-Suk Kim, Tae-Wan Kim, Chang-Goo Lee, 2002, A Study of Line Recognition and Driving Direction Control on Vision based AGV, The Korean Institute of Electrical Engineers(KIEE) Summer Conference 2002, Vol. 2002, No. 7, pp. 2341-2343Google Search
7 
Man-Hee Lee, Whang Cho, 2005, Indoor Environment Recognition Method for Indoor Autonomous Mobile Robot, The Transaction of The Korean Institute of Electrical Engineers D, Vol. 54, No. 6, pp. 366-371Google Search
8 
Mi-Ran Gu, Kang-Seok Lee, Dae-Seong Kang, 2010, Image Noise Reduction using Modified Gaussian Filter by Estimated Standard Deviation of Noise, The Journal of Korean Institute of Information Technology, Vol. 8, No. 12, pp. 111-117Google Search
9 
Hye-Kyoung Jang, 2012, The more environmentally robust edge detection of moving objects using improved Canny edge detector and Freeman chain code, The Journal of the KICS, Vol. 37, No. 2, pp. 37-42Google Search
10 
Jeong-Su Oh, 2018, Straight Line Detection Using PCA and Hough Transform, Journal of the Korea Institute of Information and Communication Engineering, Vol. 22, No. 2, pp. 227-232DOI

์ €์ž์†Œ๊ฐœ

๊น€์ •์ฃผ (Jung-Ju Kim)
../../Resources/kiee/KIEE.2021.70.6.945/au1.png

He received his B.S in Electronics Engineering, M.S. degree and PH.D. Cand. in Electronics Engineering from Hoseo University, Asan, Korea.

He is currently Professor in Robotics and Automation Engineering at Hoseo University, Asan, Korea since 2021.

His research interests are pattern recognition, image processing, deep learnig, embedded system application, and automotive mobile robot.

๊น€๋™์ง„ (Dong-Jin Kim)
../../Resources/kiee/KIEE.2021.70.6.945/au2.png

He received his B.S. in Control and Instrumentation Engineering, and M.S. and PH.D. degree in Electronics Engineering from Hoseo University, Asan, Korea, in 1998, 2000 and 2007, respectively.

He is currently Professor in Robotics and Automation Engineering at Hoseo University, Asan, Korea since 2007.

His research interests sensors, embedded system application, and smart factories.

๊ตฌ๊ฒฝ์™„ (Kyung-Wan Koo)
../../Resources/kiee/KIEE.2021.70.6.945/au3.png

1983 Chungnam National University, Department of Electronics Engineering.

1992 Chungnam National University, Department of Electronics Engineering (Ph.D.)

1987~1989 Hyundai Electronics Senior Researcher. 1989~1994 Assistant Professor, Dept. of Electronics, Chungcheong Junior College.

1994~2005 Associate Professor, School of Electronic and Information Engineering, Yeungdong University.

2005~present Professor, Department of Automotive ICT Engineering, Hoseo University.

E-mail: alarmkoo@hoseo.edu