[Computer Vision/OpenCV] 18. Edge Detection(2) - Canny Edge Detector

2025. 4. 23. 13:39ยท๐Ÿฆ„AI/Computer Vision
๋ชฉ์ฐจ
  1. 1. Edge Detection์˜ Criteria
  2. 2. Canny Edge Detector
  3. Step 1)  Image Filtering
  4. Step 2) Non-maximum Suppression
  5. Step 3) Double Thresholding
  6. Tutorial

์˜ค๋Š˜๋‚  ๊ฐ€์žฅ ๋„๋ฆฌ ์‚ฌ์šฉ๋˜๋Š” ์—ฃ์ง€ ๊ฒ€์ถœ ๊ธฐ๋ฒ• ์ค‘ ํ•˜๋‚˜์ธ Canny Edge Detection์— ๋Œ€ํ•ด์„œ ์•Œ์•„๋ณด๋„๋ก ํ•˜์ž.

 

Canny Edge Detector๋Š” ์˜์ƒ ๋‚ด ์˜๋ฏธ ์žˆ๋Š” ๊ฒฝ๊ณ„๋ฅผ ์•ˆ์ •์ ์œผ๋กœ ๊ฒ€์ถœํ•˜๊ธฐ ์œ„ํ•œ ๊ณ ์ „์ ์ธ ์•Œ๊ณ ๋ฆฌ์ฆ˜์ด๋‹ค.

์ด ๋ฐฉ์‹์€ ๊ธฐ์กด์— ์•Œ์•„๋ดค๋˜ ๋ฐฉ์‹๋“ค ์ฒ˜๋Ÿผ gradient๋ฅผ ๊ณ„์‚ฐํ•˜๋Š” ๊ฒƒ์— ๊ทธ์น˜์ง€ ์•Š๊ณ , ๋‹ค์–‘ํ•œ ํ›„์ฒ˜๋ฆฌ ๋‹จ๊ณ„๋ฅผ ์ถ”๊ฐ€ํ•ด์„œ

๋…ธ์ด์ฆˆ์— ๊ฐ•ํ•˜๋ฉด์„œ๋„ ์–‡๊ณ  ๋ช…ํ™•ํ•œ ์—ฃ์ง€๋ฅผ ์ถ”์ถœํ•  ์ˆ˜ ์žˆ๋„๋ก ์„ค๊ณ„๋˜์—ˆ๋‹ค.

 

1. Edge Detection์˜ Criteria

์ด์ƒ์ ์ธ ์—ฃ์ง€ ๊ฒ€์ถœ ๊ธฐ์ค€์€ ํฌ๊ฒŒ ์„ธ ๊ฐ€์ง€๋กœ ์š”์•ฝ๋œ๋‹ค.

 

์ฒซ์งธ, ๊ฒ€์ถœ ์˜ค์ฐจ์œจ์ด ๋‚ฎ์•„์•ผ ํ•œ๋‹ค. 

์ด๋Š” ๊ฐ€๋Šฅํ•œ ๋ชจ๋“  ์ง„์งœ ์—ฃ์ง€๋ฅผ ๋†“์น˜์ง€ ์•Š๊ณ  ํƒ์ง€ํ•ด์•ผ ํ•œ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค.

 

๋‘˜์งธ, ๊ฒ€์ถœ๋œ ์—ฃ์ง€๋Š” ์‹ค์ œ ์—ฃ์ง€์˜ ์œ„์น˜์™€ ์ตœ๋Œ€ํ•œ ๊ทผ์ ‘ํ•ด์•ผ ํ•œ๋‹ค.

์ฆ‰, Localization์ด ์ž˜ ๋˜์–ด์•ผ ํ•œ๋‹ค. ์ด๋ฅผ ํ†ตํ•ด ๊ฒฝ๊ณ„์˜ ์ •ํ™•ํ•œ ์œ„์น˜๋ฅผ ๋ณด์žฅํ•  ์ˆ˜ ์žˆ๋‹ค.

 

๋งˆ์ง€๋ง‰์œผ๋กœ, ํ•˜๋‚˜์˜ ์—ฃ์ง€์— ๋Œ€ํ•ด ๋‹ค์ˆ˜์˜ ์‘๋‹ต์ด ์•„๋‹Œ ๋‹จ์ผ ํ”ฝ์…€(single response)๋กœ ํ‘œํ˜„๋˜์–ด์•ผ ํ•œ๋‹ค.

์ด๋Š” ๊ฒฐ๊ณผ์ ์œผ๋กœ ์žก์Œ์ด๋‚˜ ์ค‘๋ณต ์—ฃ์ง€๋ฅผ ์ค„์ด๋Š” ๋ฐ ๊ธฐ์—ฌํ•œ๋‹ค.

 

์ด๋Ÿฐ ๊ธฐ์ค€์„ ๋งŒ์กฑํ•˜๊ธฐ ์œ„ํ•ด์„œ Canny Edge Detector๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™์€ ๊ณผ์ •์„ ํฌํ•จํ•˜๊ฒŒ ๋œ๋‹ค.

edge detectionํ›„ non-maximum suppression๊ณผ double thresholding๊ณผ์ •์„ ์ˆ˜ํ–‰ํ•œ๋‹ค.

 

2. Canny Edge Detector

์œ„์—์„œ ์ œ์‹œ๋œ Edge Detection์˜ ๊ธฐ์ค€์„ ๋งŒ์กฑ์‹œํ‚ค๊ธฐ ์œ„ํ•ด์„œ

Canny Edge Detector๊ฐ€ ์–ด๋–ค ๊ณผ์ •์„ ๊ฑฐ์น˜๋Š”์ง€ ์กฐ๊ธˆ ๋” ์ž์„ธํžˆ, ์ˆœ์„œ๋Œ€๋กœ ์‚ดํŽด๋ณด์ž.

 

Step 1)  Image Filtering

์•ž์„  ํฌ์ŠคํŒ…์—์„œ ๋ดค๋˜ Low-pass Filter์™€ High-pass Filter๋ฅผ ์‚ฌ์šฉํ•ด์„œ ์—ฃ์ง€๋ฅผ ๊ฒ€์ถœํ•˜๋Š” ๊ณผ์ •์ด๋‹ค.

 

์ฆ‰, Gaussian Filter์™€ Sobel Filter๋ฅผ ์‚ฌ์šฉํ•ด์„œ(DoG) ์—ฃ์ง€๋ฅผ ๊ฒ€์ถœํ•œ๋‹ค. 

์ด ๊ณผ์ •์„ ํ†ตํ•ด์„œ ์ด๋ฏธ์ง€์˜ edge์™€, edge ์˜ ๊ฐ๋„ ์ •๋ณด๋ฅผ ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

 

๋‹ค์Œ์ฒ˜๋Ÿผ Sobel Filtering ์ง„ํ–‰ํ•œ๋‹ค.
Edge Magmitude and Angle

์—ฃ์ง€์˜ ํฌ๊ธฐ์™€ ๋ฐฉํ–ฅ์ •๋ณด๋Š” ์ดํ›„์˜ Non-Maximum Suppression๊ณผ Double Thresholding์— ์‚ฌ์šฉ๋œ๋‹ค.

Step 2) Non-maximum Suppression

Gradient ๊ณ„์‚ฐ์„ ํ†ตํ•ด ์–ป์–ด์ง„ ์—ฃ์ง€ ํ›„๋ณด๋Š” ๋‘๊ป๊ณ  ๋ถˆ๋ช…ํ™•ํ•œ ๊ฒฝ๊ณ„์„ ์„ ํฌํ•จํ•  ์ˆ˜ ์žˆ๋‹ค.

์ด์ „ ํฌ์ŠคํŒ… https://he-kate1130.tistory.com/145 ์—์„œ ์‚ดํŽด๋ณด์•˜๋“ฏ์ด

์Šค๋ฌด๋”ฉ์„ ์ ์šฉํ•˜๊ณ  ๋ฏธ๋ถ„ํ•˜์—ฌ edge๋ฅผ ๊ฒ€์ถœํ•˜๋ฉด Edge์˜ ๋ฒ”์œ„๊ฐ€ ๋„“๊ฒŒ ๊ฒ€์ถœ๋˜์–ด Localization์— ๋ฌธ์ œ๊ฐ€ ์ƒ๊ธด๋‹ค.

Edge Detection - Smoothing Tradeoff

 

์ด์ฒ˜๋Ÿผ ๊ฒ€์ถœ๋œ Gradient Magnitude๊ฐ€ ์ผ์ • ๋ฒ”์œ„ ๋‚ด์—์„œ ๋„“๊ฒŒ ๋ถ„ํฌ๋˜์–ด ์žˆ์„ ๊ฒฝ์šฐ, ํ•˜๋‚˜์˜ ์—ฃ์ง€๊ฐ€ ์—ฌ๋Ÿฌ ํ”ฝ์…€์— ๊ฑธ์ณ ํ‘œํ˜„๋˜๋Š” ํ˜„์ƒ์ด ๋ฐœ์ƒํ•œ๋‹ค.

์ด๋Ÿฌํ•œ ์ค‘๋ณต ํ‘œํ˜„์€ ์—ฃ์ง€ ๊ฒ€์ถœ ๊ฒฐ๊ณผ์˜ ์ •๋ฐ€๋„๋ฅผ ๋–จ์–ด๋œจ๋ฆฌ๋ฏ€๋กœ

์ด๋ฅผ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด Non-Maximum Suppression(NMS) ๊ธฐ๋ฒ•์ด ์ ์šฉ๋œ๋‹ค.

 

Non-Maximum Suppression์˜ ํ•ต์‹ฌ ๋ชฉ์ ์€ Gradient Angle์„ ๊ธฐ์ค€์œผ๋กœ ์ฃผ๋ณ€ ํ”ฝ์…€๋“ค๊ณผ์˜ ๋น„๊ต๋ฅผ ํ†ตํ•ด ๊ฐ€์žฅ ๊ฐ•ํ•œ ์‘๋‹ต์„ ๊ฐ€์ง€๋Š” ๋‹จ ํ•œ๊ฐœ์˜ ํ”ฝ์…€๋งŒ์„ ๋‚จ๊ธฐ๊ณ , ๋‚˜๋จธ์ง€๋Š” ์ œ๊ฑฐํ•จ์œผ๋กœ์จ ์—ฃ์ง€๋ฅผ ๊ฐ€๋Šฅํ•œ ํ•œ ์–‡๊ฒŒ ์œ ์ง€ํ•˜๋Š” ๊ฒƒ์ด๋‹ค.

 

์ด ๊ณผ์ •์€ ๋‹ค์Œ๊ณผ ๊ฐ™์€ ์ˆœ์„œ๋กœ ์ง„ํ–‰๋œ๋‹ค.

 

Edge Magmitude and Angle

  1. ์ž‘์€ window (์˜ˆ๋ฅผ ๋“ค์–ด 3x3) ์„ ๋งŒ๋“ ๋‹ค.
  2. ์ด window์˜ ์ค‘์‹ฌ์„ (x,y)์— ๋‘”๋‹ค.
  3. A(x,y)์˜ ๋ฐฉํ–ฅ์œผ๋กœ ์ด์›ƒ ํ”ฝ์…€์„ ์ฐพ๋Š”๋‹ค.
  4. ์ฐพ์€ ์ด์›ƒํ”ฝ์…€ ๊ฐ„ Gradient์˜ Magnitude(M)๋ฅผ ๋น„๊ตํ•œ๋‹ค.
  5. ์ด ์ค‘์—์„œ ๊ฐ€์žฅ ๋†’์€ ๊ฐ’๋งŒ edge๋กœ ์„ ํƒํ•œ๋‹ค.

(x,y)๋ฅผ ๊ธฐ์ค€์œผ๋กœ A(x,y)๋ฐฉํ–ฅ์˜ ์ด์›ƒํ”ฝ์…€์—์„œ M๊ฐ’์„ ๋น„๊ตํ•œ๋‹ค.

 

์—ฌ๊ธฐ์„œ ๋ฌธ์ œ๊ฐ€ ํ•˜๋‚˜ ๋” ์ƒ๊ธด๋‹ค. 

์˜์ƒ์€ ์ด์‚ฐ ์‹œ๊ทธ๋„๋กœ ๋˜์–ด์žˆ์œผ๋‹ˆ

A(x,y)๋ฐฉํ–ฅ์— ์ผ์น˜๋˜๋Š” ์ด์›ƒ ํ”ฝ์…€์ด ์กด์žฌํ•˜์ง€ ์•Š๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋Œ€๋‹ค์ˆ˜์ด๋‹ค... (์™ผ์ชฝ ์‚ฌ์ง„)

์ด ๋ฌธ์ œ ๋•Œ๋ฌธ์— A(x,y) ๊ฐ’์„ Quantizationํ•ด์„œ ์‚ฌ์šฉํ•œ๋‹ค. (์˜ค๋ฅธ์ชฝ ์‚ฌ์ง„)

 

์ด๋ ‡๊ฒŒ ํ•ด์„œ Edge์˜ ๋„ˆ๋น„๋ฅผ 1๊ฐœ์˜ ํ”ฝ์…€๋กœ ์„ ํƒํ•˜์˜€๋‹ค.

Non-max Suppression ์ ์šฉ ์ „, ํ›„

 

Step 3) Double Thresholding

์œ„์˜ ๊ณผ์ •์„ ํ†ตํ•ด์„œ ์–‡์€ ์—ฃ์ง€ ํ›„๋ณด๋“ค์ด ์„ ํƒ๋˜์—ˆ๋Š”๋ฐ, ์—ฌ์ „ํžˆ ์ด ์ค‘์˜ ์ผ๋ถ€๋Š” ์‹ค์ œ๋กœ๋Š” Edge๊ฐ€ ์•„๋‹ ์ˆ˜ ์žˆ๋‹ค.

๋”ฐ๋ผ์„œ Canny ์•Œ๊ณ ๋ฆฌ์ฆ˜์˜ ๋งˆ์ง€๋ง‰ ๋‹จ๊ณ„์—์„œ๋Š” ์ด๋Ÿฌํ•œ ์—ฃ์ง€ ํ›„๋ณด๋“ค ์ค‘ ์‹ค์ œ๋กœ ์˜๋ฏธ ์žˆ๋Š” ์—ฃ์ง€๋ฅผ ์ตœ์ข…์ ์œผ๋กœ ํ™•์ •ํ•˜๋Š” ์ ˆ์ฐจ๊ฐ€ ํ•„์š”ํ•˜๋‹ค.

์ด๋ฅผ ์œ„ํ•ด ์‚ฌ์šฉ๋˜๋Š” ๊ฒƒ์ด ๋ฐ”๋กœ Double Thresholding๊ณผ Hysteresis ๊ธฐ๋ฒ•์ด๋‹ค.

 

Double Thresholding์€ gradient magnitude์— ๊ธฐ๋ฐ˜ํ•˜์—ฌ ํ”ฝ์…€์„ ์„ธ ๊ฐ€์ง€ ๋ฒ”์ฃผ๋กœ ๋‚˜๋ˆ„๋Š” ๋ฐฉ์‹์ด๋‹ค.

๋†’์€ ์ž„๊ณ„๊ฐ’ T_H๋ณด๋‹ค ํฐ gradient ๊ฐ’(M)์„ ๊ฐ€์ง„ ํ”ฝ์…€์€ ๊ฐ•ํ•œ ์—ฃ์ง€(strong edge)๋กœ ๊ฐ„์ฃผ๋˜๊ณ , ํ™•์‹คํ•œ ์—ฃ์ง€๋กœ ๋ฐ”๋กœ ์ฑ„ํƒ๋œ๋‹ค.

๋‚ฎ์€ ์ž„๊ณ„๊ฐ’ T_L๋ณด๋‹ค ์ž‘์€ M๊ฐ’์„ ๊ฐ€์ง€๋Š” ํ”ฝ์…€์€ ์—ฃ์ง€๊ฐ€ ์•„๋‹Œ ๊ฒƒ์œผ๋กœ ํŒ๋‹จ๋˜์–ด ์ฆ‰์‹œ ์ œ๊ฑฐ๋œ๋‹ค.

 

๋ฌธ์ œ๋Š” ์ค‘๊ฐ„ ๋ฒ”์œ„โ€‹์— ํ•ด๋‹นํ•˜๋Š” M ๊ฐ’์„ ๊ฐ€์ง„ ํ”ฝ์…€๋“ค์ธ๋ฐ, ์ด๋“ค์€ gradient ํฌ๊ธฐ๋งŒ์œผ๋กœ๋Š” ํ™•์ •์ ์ธ ํŒ๋‹จ์ด ์–ด๋ ต๊ธฐ ๋•Œ๋ฌธ์— ์ฃผ๋ณ€ ๋งฅ๋ฝ์— ๋”ฐ๋ผ ์ฒ˜๋ฆฌ๋œ๋‹ค.

์ฃผ๋ณ€ ๋งฅ๋ฝ์— ๋”ฐ๋ผ ์ฒ˜๋ฆฌํ•œ๋‹ค๋Š” ์˜๋ฏธ๋Š” ํžˆ์Šคํ…Œ๋ฆฌ์‹œ์Šค(Hysteresis)๋ฅผ ์ ์šฉํ•œ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค.

 

ํžˆ์Šคํ…Œ๋ฆฌ์‹œ์Šค๋Š” Edge Connectivity ํŒ๋‹จ์— ๊ธฐ๋ฐ˜ํ•œ ๋ณด์™„ ๊ธฐ๋ฒ•์œผ๋กœ, ์•ฝํ•œ ์—ฃ์ง€ ํ”ฝ์…€์ด ์ฃผ๋ณ€์— ๊ฐ•ํ•œ ์—ฃ์ง€ ํ”ฝ์…€๊ณผ ์—ฐ๊ฒฐ๋˜์–ด ์žˆ์„ ๊ฒฝ์šฐ, ์ด๋ฅผ ์œ ํšจํ•œ ์—ฃ์ง€๋กœ ์ธ์ •ํ•œ๋‹ค.

๋ฐ˜๋Œ€๋กœ ์ฃผ๋ณ€์— ๊ฐ•ํ•œ ์—ฃ์ง€๊ฐ€ ์—†์œผ๋ฉด, ํ•ด๋‹น ํ”ฝ์…€์€ ์žก์Œ์œผ๋กœ ํŒ๋‹จ๋˜์–ด ์ œ๊ฑฐ๋œ๋‹ค.

์ด์™€ ๊ฐ™์€ ์ฒ˜๋ฆฌ ๊ณผ์ •์„ ํ†ตํ•ด ์ง„์งœ ์—ฃ์ง€๋ฅผ ๋†“์น˜์ง€ ์•Š์œผ๋ฉด์„œ๋„ ๋…ธ์ด์ฆˆ์— ๊ธฐ์ธํ•œ ์ž˜๋ชป๋œ ์—ฃ์ง€๋Š” ํšจ๊ณผ์ ์œผ๋กœ ์ œ๊ฑฐํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋œ๋‹ค.

 

์ฆ‰, ์„ค์ •ํ•œ ๋‘ Threshold์‚ฌ์ด์˜ M๊ฐ’์„ ๊ฐ€์ง„ ํ”ฝ์…€๋“ค์€ ์ด์›ƒ์ด Edge๋ผ๋ฉด ๋™์ผํ•˜๊ฒŒ Edge ๋กœ ์ฑ„ํƒ๋œ๋‹ค.

 

์ด ๋‹จ๊ณ„๋Š” Canny ์•Œ๊ณ ๋ฆฌ์ฆ˜์ด ์ถ”๊ตฌํ•˜๋Š” Low Error Rate์™€ Single Response ๊ธฐ์ค€์„ ๋งŒ์กฑ์‹œํ‚ค๋Š” ๋ฐ ํ•ต์‹ฌ์ ์ธ ์—ญํ• ์„ ํ•œ๋‹ค. ํžˆ์Šคํ…Œ๋ฆฌ์‹œ์Šค๋ฅผ ํ†ตํ•œ ์—ฐ๊ฒฐ์„ฑ ๊ธฐ๋ฐ˜ ํŒ๋‹จ์€ ์•ฝํ•œ ์—ฃ์ง€๋ฅผ ์˜๋ฏธ ์žˆ๋Š” ๊ตฌ์กฐ๋กœ ํ†ตํ•ฉํ•จ์œผ๋กœ์จ ์ง„์งœ ์—ฃ์ง€๋ฅผ ์ตœ๋Œ€ํ•œ ๋†“์น˜์ง€ ์•Š๋„๋ก ๋ณด์™„(Low Error Rate)ํ•˜์—ฌ ์ตœ์ข… ์—ฃ์ง€ ๊ฒฐ๊ณผ๋ฅผ ๋ช…ํ™•ํ•˜๊ฒŒ ์ •๋ฆฌํ•œ๋‹ค.

 

Double Thresholding์ ์šฉ ์ „, ํ›„

 

Tutorial

ํ•ด๋‹น ํŠœํ† ๋ฆฌ์–ผ๊ณผ ๋™์ผํ•˜๊ฒŒ ์ง„ํ–‰ํ•œ๋‹ค.

https://docs.opencv.org/4.11.0/da/d22/tutorial_py_canny.html

 

OpenCV: Canny Edge Detection

Goal In this chapter, we will learn about Concept of Canny edge detection OpenCV functions for that : cv.Canny() Theory Canny Edge Detection is a popular edge detection algorithm. It was developed by John F. Canny in It is a multi-stage algorithm and we wi

docs.opencv.org

 

์‹คํ–‰ ๊ฒฐ๊ณผ

 

 

#include <opencv2/opencv.hpp>
#include <iostream>
using namespace cv;


int main() {
    // ์ด๋ฏธ์ง€ ์ฝ๊ธฐ (Grayscale)
    Mat img = imread("bridge.jpg", IMREAD_GRAYSCALE);
    if (img.empty()) {
        std::cerr << "Image load failed!" << std::endl;
        return -1;
    }

    // ๊ฒฐ๊ณผ ์ €์žฅ ํด๋”
    std::string output_dir = "./output/";
    imwrite(output_dir + "original.png", img);

    Mat canny_edges;
    Canny(img, canny_edges, 100, 200);

    imwrite(output_dir + "canny.png", canny_edges);

    std::cout << "All processing done. Check output folder." << std::endl;
    return 0;
}

https://github.com/mingyung-park/CV_Study/tree/main/07_EdgeCornerDetection

 

CV_Study/07_EdgeCornerDetection at main ยท mingyung-park/CV_Study

OpenCV์™€ C++, Python์„ ํ™œ์šฉํ•œ ์ปดํ“จํ„ฐ ๋น„์ „ ์ด๋ก  ๋ฐ ๊ตฌํ˜„ ์ •๋ฆฌ ๋ ˆํฌ์ง€ํ† ๋ฆฌ์ž…๋‹ˆ๋‹ค. ๊ธฐ์ดˆ ์ด๋ก ๋ถ€ํ„ฐ ์‹ค์Šต ์ฝ”๋“œ๊นŒ์ง€ ๋‹จ๊ณ„๋ณ„๋กœ ํ•™์Šตํ•˜๋ฉฐ ์ •๋ฆฌํ•ฉ๋‹ˆ๋‹ค. - mingyung-park/CV_Study

github.com

 

'๐Ÿฆ„AI > Computer Vision' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๋‹ค๋ฅธ ๊ธ€

[Computer Vision/OpenCV] 19. Corner Detection - Harris Corner Detection  (0) 2025.04.24
[Computer Vision/OpenCV] 17. Edge Detection(1) - Sobel, Laplacian of Gaussian Filter  (0) 2025.04.22
[Computer Vision/OpenCV] 16. Edge Detection ๊ณผ Smoothing Tradeoff  (0) 2025.04.10
[Computer Vision/OpenCV] 15. Segmentation  (1) 2025.04.06
[Computer Vision/OpenCV] 14. Image Denoising  (0) 2025.04.05
  1. 1. Edge Detection์˜ Criteria
  2. 2. Canny Edge Detector
  3. Step 1)  Image Filtering
  4. Step 2) Non-maximum Suppression
  5. Step 3) Double Thresholding
  6. Tutorial
'๐Ÿฆ„AI/Computer Vision' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๋‹ค๋ฅธ ๊ธ€
  • [Computer Vision/OpenCV] 19. Corner Detection - Harris Corner Detection
  • [Computer Vision/OpenCV] 17. Edge Detection(1) - Sobel, Laplacian of Gaussian Filter
  • [Computer Vision/OpenCV] 16. Edge Detection ๊ณผ Smoothing Tradeoff
  • [Computer Vision/OpenCV] 15. Segmentation
mingyung
mingyung
  • mingyung
    KATE.log
    mingyung
  • ์ „์ฒด
    ์˜ค๋Š˜
    ์–ด์ œ
    • ๋ถ„๋ฅ˜ ์ „์ฒด๋ณด๊ธฐ
      • ๐Ÿฆ„AI
        • Computer Vision
        • DL
        • ML
        • Paper
        • NLP
        • Reinforcement Learning
      • ๐ŸAlgorithm
      • ๐Ÿฅ”Project
        • Reversi Agent
        • Team22
        • [Project] Threepark
      • Computer Science
        • SW Engineering
        • Database
        • Computer Network
        • Compiler
        • Data structures
      • ์žก๋™์‚ฌ๋‹ˆ
      • Dev
        • Spring
  • ์ตœ๊ทผ ๊ธ€

  • ์ธ๊ธฐ ๊ธ€

  • ํƒœ๊ทธ

    CV
    Computer Vision
    computer network
    Filtering
    ์ธํ„ฐ๋„ท
    loss
    Gaussian
    network
    ์ž๋ฃŒ๊ตฌ์กฐ
    ML
    Edge Detection
    ์ปดํ“จํ„ฐ ๋„คํŠธ์›Œํฌ
    OSP
    Loss function
    localization
    simple thresholding
    optimum global thresholding
    Image Processing
    Internet
    harris corner
  • ์ตœ๊ทผ ๋Œ“๊ธ€

  • hELLOยท Designed By์ •์ƒ์šฐ.v4.10.3
mingyung
[Computer Vision/OpenCV] 18. Edge Detection(2) - Canny Edge Detector
์ƒ๋‹จ์œผ๋กœ

ํ‹ฐ์Šคํ† ๋ฆฌํˆด๋ฐ”

๋‹จ์ถ•ํ‚ค

๋‚ด ๋ธ”๋กœ๊ทธ

๋‚ด ๋ธ”๋กœ๊ทธ - ๊ด€๋ฆฌ์ž ํ™ˆ ์ „ํ™˜
Q
Q
์ƒˆ ๊ธ€ ์“ฐ๊ธฐ
W
W

๋ธ”๋กœ๊ทธ ๊ฒŒ์‹œ๊ธ€

๊ธ€ ์ˆ˜์ • (๊ถŒํ•œ ์žˆ๋Š” ๊ฒฝ์šฐ)
E
E
๋Œ“๊ธ€ ์˜์—ญ์œผ๋กœ ์ด๋™
C
C

๋ชจ๋“  ์˜์—ญ

์ด ํŽ˜์ด์ง€์˜ URL ๋ณต์‚ฌ
S
S
๋งจ ์œ„๋กœ ์ด๋™
T
T
ํ‹ฐ์Šคํ† ๋ฆฌ ํ™ˆ ์ด๋™
H
H
๋‹จ์ถ•ํ‚ค ์•ˆ๋‚ด
Shift + /
โ‡ง + /

* ๋‹จ์ถ•ํ‚ค๋Š” ํ•œ๊ธ€/์˜๋ฌธ ๋Œ€์†Œ๋ฌธ์ž๋กœ ์ด์šฉ ๊ฐ€๋Šฅํ•˜๋ฉฐ, ํ‹ฐ์Šคํ† ๋ฆฌ ๊ธฐ๋ณธ ๋„๋ฉ”์ธ์—์„œ๋งŒ ๋™์ž‘ํ•ฉ๋‹ˆ๋‹ค.