Presentasi sedang didownload. Silahkan tunggu

Presentasi sedang didownload. Silahkan tunggu

Edge Detection Dr. Aniati Murni (R 1202) Dina Chahyati, SKom (R 1226)

Presentasi serupa


Presentasi berjudul: "Edge Detection Dr. Aniati Murni (R 1202) Dina Chahyati, SKom (R 1226)"— Transcript presentasi:

1 Edge Detection Dr. Aniati Murni (R 1202) Dina Chahyati, SKom (R 1226)
Fakultas Ilmu Komputer Universitas Indonesia ©AniatiMurni

2 Gradient Brightness gradient of image f(x,y): Digital derivative:
umumnya n=1.

3 Magnitude of gradient vector
Rumus 1: Rumus 2: Rumus 3: The quickest speed with which the intensity changes at f(x,y)

4 Direction of gradient vector (1)
The direction in which the intensity changes the quickest at f(x,y) Direction

5 Direction of gradient vector (2)
Edge contour direction: along the contour, right side is white (high value) Edge gradient direction: orthogonal to the contour, towards white (high value)

6 Direction of gradient vector (3)
Sumber: MSU

7 1st derivative and 2nd derivative
f(I,j-1) f(I-1,j) f(I,j) f(I+1,j) f(I,j+1)

8 Laplacian Operator (1) Citra Kontinue: Citra Dijital:

9 Laplacian Operator (2) gaverage menguatkan respon frekwensi rendah dan melemahkan respon frekwensi tinggi  (2g-gaverage) akan menguatkan respon frekwensi tinggi relatif terhadap frekwensi rendah.

10 Konsep Zero-Crossing 1-D image 1st derivative 2nd derivative
Frekwensi rendah dan frekwensi tinggi. (a) Perubahan intensitas; (b) Mempunyai peak; (c) Steep zero-crossing. Sumber: MSU

11 Contoh Kernel Edge Detector (1)
Gerald K. Moore: directional edge detection E-W N-S NW-SE NE-SW WNW-ESE NNW-SSE ENE-WSW NNE-SSW

12 Contoh Kernel Edge Detector (2)
Baxter: directional edge detection Utara U-T Timur S-T Selatan S-B Barat U-B

13 Contoh Kernel Edge Detector (2)
Robert (1962): Prewitt (1970): Sobel (1970):

14 Contoh Kernel Edge Detector (3)
Kirsh (1977): n=1/ n=1 n=2

15 Contoh Kernel Edge Detector (4)
Robinson (1977) Frei-Chen (1977)

16 1st derivative and 2nd derivative
Contoh image: Hasil 1st derivative (outlining): Hasil 2nd derivative (retaining original image):

17 Laplacian of Gaussion Filtering (1)
Gaussian operator (LPF): Gaussian blurring adalah Smoothing untuk menghilangkan noise, dengan nilai yang besar atau yang kecil 1-D: 2-D:

18 Laplacian of Gaussion Filtering (2)
Laplacian operator (HPF): Laplacian bertujuan untuk meningkatkan kwalitas detil (detail enhancement) Laplacian of Gaussian filtering bertujuan untuk menghilangkan noise dan meningkatkan kwalitas detil.

19 Laplacian of Gaussion Filtering (3)
Laplacian of Gaussian: dengan Selanjutnya dicari lokasi zero-crossing untuk menentukan garis batas antara hitam dan putih.

20 Laplacian of Gaussion Filtering (4)
(a) (b) (c) (a) Original image (320 x 320 pixels) (b) Gaussian filtering dengan = 8 piksel (Sumber: MSU) (c) Gaussian filtering dengan = 4 piksel

21 Laplacian of Gaussion Filtering (5)
(a) (b) © (a) Laplacian of Gaussian (b) Positive = putih dan negative = hitam (c) zero-crossings (Sumber: MSU)


Download ppt "Edge Detection Dr. Aniati Murni (R 1202) Dina Chahyati, SKom (R 1226)"

Presentasi serupa


Iklan oleh Google