Edge Detection Speaker: Che-Ming Hu Advisor: Jian-Jiun Ding

Slides:



Advertisements
Similar presentations
Spatial Filtering (Chapter 3)
Advertisements

Topic 6 - Image Filtering - I DIGITAL IMAGE PROCESSING Course 3624 Department of Physics and Astronomy Professor Bob Warwick.
Spatial Filtering.
EDGE DETECTION ARCHANA IYER AADHAR AUTHENTICATION.
DREAM PLAN IDEA IMPLEMENTATION Introduction to Image Processing Dr. Kourosh Kiani
Edge detection Goal: Identify sudden changes (discontinuities) in an image Intuitively, most semantic and shape information from the image can be encoded.
EE663 Image Processing Edge Detection 2 Dr. Samir H. Abdul-Jauwad Electrical Engineering Department King Fahd University of Petroleum & Minerals.
Median Filter If the objective is to achieve noise reduction rather than blurring, an alternative approach is to use median filters. That is, the gray.
MSU CSE 803 Stockman Linear Operations Using Masks Masks are patterns used to define the weights used in averaging the neighbors of a pixel to compute.
Edge Detection Phil Mlsna, Ph.D. Dept. of Electrical Engineering
Segmentation (Section 10.2)
Scale Invariant Feature Transform (SIFT)
Introduction to Computer Vision CS / ECE 181B Thursday, April 22, 2004  Edge detection (HO #5)  HW#3 due, next week  No office hours today.
2-D, 2nd Order Derivatives for Image Enhancement
Filters and Edges. Zebra convolved with Leopard.
CS 376b Introduction to Computer Vision 03 / 04 / 2008 Instructor: Michael Eckmann.
© 2010 Cengage Learning Engineering. All Rights Reserved.
MSU CSE 803 Linear Operations Using Masks Masks are patterns used to define the weights used in averaging the neighbors of a pixel to compute some result.
Segmentation-Based Image Compression 以影像切割為基礎的影像壓縮技術 Speaker: Jiun-De Huang Advisor: Jian-Jiun Ding Graduate Institute of Communication Engineering National.
Edge Detection Hao Huy Tran Computer Graphics and Image Processing CIS 581 – Fall 2002 Professor: Dr. Longin Jan Latecki.
Computer Vision Spring ,-685 Instructor: S. Narasimhan WH 5409 T-R 10:30 – 11:50am.
Introduction to Image Processing Grass Sky Tree ? ? Sharpening Spatial Filters.
Introduction to Image Processing
EDGE DETECTION IN COMPUTER VISION SYSTEMS PRESENTATION BY : ATUL CHOPRA JUNE EE-6358 COMPUTER VISION UNIVERSITY OF TEXAS AT ARLINGTON.
Chapter 10 Image Segmentation.
Image Processing Segmentation 1.Process of partitioning a digital image into multiple segments (sets of pixels). 2. Clustering pixels into salient image.
Chapter 10, Part I.  Segmentation subdivides an image into its constituent regions or objects.  Image segmentation methods are generally based on two.
Instructor: S. Narasimhan
Spatial Filtering.
CS654: Digital Image Analysis Lecture 24: Introduction to Image Segmentation: Edge Detection Slide credits: Derek Hoiem, Lana Lazebnik, Steve Seitz, David.
Digital Image Processing Lecture 16: Segmentation: Detection of Discontinuities Prof. Charlene Tsai.
Edge Detection and Geometric Primitive Extraction Jinxiang Chai.
Sejong Univ. Edge Detection Introduction Simple Edge Detectors First Order Derivative based Edge Detectors Compass Gradient based Edge Detectors Second.
Mestrado em Ciência de Computadores Mestrado Integrado em Engenharia de Redes e Sistemas Informáticos VC 15/16 – TP7 Spatial Filters Miguel Tavares Coimbra.
Edge-Detection and Wavelet Transform
Advanced Digital Signal Processing final report NAME : YI-WEI CHEN TEACHER : JIAN-JIUN DING.
主講者 : 陳建齊. Outline & Content 1. Introduction 2. Thresholding 3. Edge-based segmentation 4. Region-based segmentation 5. conclusion 2.
Chapter 9: Image Segmentation
Digital Image Processing Lecture 16: Segmentation: Detection of Discontinuities May 2, 2005 Prof. Charlene Tsai.
Announcements Project 0 due tomorrow night. Edge Detection Today’s readings Cipolla and Gee (handout) –supplemental: Forsyth, chapter 9Forsyth For Friday.
Course 5 Edge Detection. Image Features: local, meaningful, detectable parts of an image. edge corner texture … Edges: Edges points, or simply edges,
Lecture 04 Edge Detection Lecture 04 Edge Detection Mata kuliah: T Computer Vision Tahun: 2010.
Machine Vision Edge Detection Techniques ENT 273 Lecture 6 Hema C.R.
Computer Vision Image Features Instructor: Dr. Sherif Sami Lecture 4.
Instructor: Mircea Nicolescu Lecture 7
Digital Image Processing Week V Thurdsak LEAUHATONG.
Digital Image Processing CSC331
Sliding Window Filters Longin Jan Latecki October 9, 2002.
1 Edge Operators a kind of filtering that leads to useful features.
Spatial Filtering (Chapter 3) CS474/674 - Prof. Bebis.
Image Enhancement in the Spatial Domain.
EDGE DETECTION Dr. Amnach Khawne. Basic concept An edge in an image is defined as a position where a significant change in gray-level values occur. An.
Miguel Tavares Coimbra
Edge Detection slides taken and adapted from public websites:
Image Filtering Spatial filtering
Edge Detection Phil Mlsna, Ph.D. Dept. of Electrical Engineering Northern Arizona University.
Chapter 10 Image Segmentation
Digital Image Processing CSC331
Digital Image Processing Lecture 16: Segmentation: Detection of Discontinuities Prof. Charlene Tsai.
Image Pre-Processing in the Spatial and Frequent Domain
Edge detection Goal: Identify sudden changes (discontinuities) in an image Intuitively, most semantic and shape information from the image can be encoded.
Contourlet Transforms For Feature Detection
a kind of filtering that leads to useful features
Edge Detection CSE 455 Linda Shapiro.
a kind of filtering that leads to useful features
Linear Operations Using Masks
Feature Detection .
IT472 Digital Image Processing
IT472 Digital Image Processing
Introduction to Artificial Intelligence Lecture 22: Computer Vision II
Presentation transcript:

Edge Detection Speaker: Che-Ming Hu Advisor: Jian-Jiun Ding Graduate Institute of Communication Engineering National Taiwan University, Taipei, Taiwan, ROC 2019/1/3

Outline Introduction Edge detection method Simulation results Differentiation Hibert Transform Short Response Hilbert Transform Improved Harri’s Algorithm Simulation results Conclusion Reference 2019/1/3

Introduction Fig.1. Edges are boundaries between different textures.Edge also can be defined as discontinuities inimage intensity from one pixel to another.. [4] 2019/1/3

EDGE DETECTION METHOD Fist-Order Derivative Edge Detection Second-Order Derivative Edge Detection Hilbert Transform for Edge Detection Short Response Hilbert Transform for Edge Detection Improved Harri’s Algorithm For Corner And Egde Detections 2019/1/3

Fist-Order Derivative Edge Detection Introduction The Roberts operators The Prewitt operators The Sobel operators First-Order of Gausssian (FDOG) 2019/1/3

Fist-Order Derivative Edge Detection Definition: the gradient vector the magnitude of this vector the direction of the gradient vector 2019/1/3

Fist-Order Derivative Edge Detection Fig.2. Orthogonal gradient generation.[2] The gradient along the line normal to the edge slope The spatial gradient amplitude The gradient amplitude combination 2019/1/3

The Roberts operators 2019/1/3

The Prewitt operators 2019/1/3

The Sobel operators 2019/1/3

Operators Fig.3. Impulse response arrays for 3 ×3 orthogonal differential gradient edge operators.[2] 2019/1/3

First-Order of Gausssian It is hard to find the gradient by using the equation In order to simplify the computation M=|Mx|+|My| 2019/1/3

First-order derivative edge detection Fig.4. Using 1st-order differentiation to detect (a) the sharp edges, (c) the step edges with noise, and (e) the ramp edges. (b)(d)(e) are the results of differentiation of (a)(c)(e).[3] 2019/1/3

Canny Edge Detection Good detection Good localization Single response Many edge candidate The accurate edge 2019/1/3

Fig.5. Comparison with FDOG and the Sobel operators Simulation The original image FDOG The Sobel operators Fig.5. Comparison with FDOG and the Sobel operators 2019/1/3

Fig.6. Simulation of first-order of derivative edge detection[3] (a) Original image. (b) Roberts operator. (b) (d) (c) Prewitt operator. (d) Sobel operator. Fig.6. Simulation of first-order of derivative edge detection[3] 2019/1/3

Second-order derivative edge detection The Laplacian of a 2-D function f (x, y) is a second-order derivative defined as: 2019/1/3

Second-order derivative edge detection 2019/1/3

Second-order derivative edge detection The 2-D Gaussian function: The Laplacian of Gaussian (LOG): 2019/1/3

Second-order derivative edge detection Fig.7. An 11×11 mask approximation to Laplacian of Gaussian (LOG).[3] 2019/1/3

Second-order derivative edge detection Sensitive to noise Gaussian function Low-pass filter 2019/1/3

Second-order derivative edge detection (a) Original image. (b) Laplacian mask of Fig.(a) (b) (d) (c) Laplacian mask of Fig.(b) (d) LOG mask of Fig.(b) Fig.9. Simulation of second-order of derivative edge detection[3] 2019/1/3

Second-order derivative edge detection Fig.10. Using 1st-order differentiation to detect (a) the sharp edges, (c) the step edges with noise, and (e) the ramp edges. (b)(d)(e) are the results of differentiation of (a)(c)(e).[3] 2019/1/3

Second-order derivative edge detection The Drawbacks of the Differentiation Method for Edge Detection : Sensitivity to noise Not good for ramp edges Make no difference between the significant edge and the detailed edge 2019/1/3

Hilbert Transform for Edge Detection H(f)= −jsgn(f), sgn(f)= 1 ,when f>0, sgn(f)=- 1 ,when f<0, sgn(0)= 0 2019/1/3

Hilbert Transform for Edge Detection Fig.11. Using HLTs to detect (a) the sharp edges, (c) the step edges with noise, and (e) the ramp edges. (b)(d)(e) are the results of the HLTs of (a)(c)(e). [3] 2019/1/3

Hilbert Transform for Edge Detection Slove the problem of differentiation Edge is too thick SRHLT 2019/1/3

Short Response Hilbert Transform for Edge Detection From We obtain: 2019/1/3

Short Response Hilbert Transform for Edge Detection we can define the short response Hilbert transform (SRHLT) as: SRHLT(depend on the value of b) Large (Differentiation) Small (HLT) b 2019/1/3

Short Response Hilbert Transform for Edge Detection Fig.12.Impulse responses and their FTs of the SRHLT for different b. We can compare them with the impulse response of the differential operation and the original HLT.[3] 2019/1/3

Short Response Hilbert Transform for Edge Detection Fig.13. Using SRHLTs to detect the sharp edges, the step edges with noise, and the ramp edges. Here we choose b = 1, 4, 12, and 30. [3] 2019/1/3

Short Response Hilbert Transform for Edge Detection Without noise: (a) Original image (b) Results of differentiation (c) Results of the HLT (d) Results of the SRHLT, b=8 Fig.14. Experiments that use differentiation, the HLT, and the SRHLT (b=8) to do edge detection for Lena image.[3] 2019/1/3

Short Response Hilbert Transform for Edge Detection With noise: (a) Lena + noise, SNR = 32 (b) Results of differentiation (c) Results of the HLT (d) Results of the SRHLT, b=8 Fig.15. Experiments that use differentiation, the HLT, and the SRHLT (b=8) to detect the edges of Lena image interfered by noise.[3] 2019/1/3

Short Response Hilbert Transform for Edge Detection (a) b = 0.5 for Lena image (b) b = 0.5 for Lena + noise (c) b = 8 for Lena image (d) b = 8 for Lena + noise (e) b = 20 for Lena image (f) b = 20 for Lena + noise Fig.16. Experiment of SRHLT with different b (0.5, 8, and 20)[3] 2019/1/3

Short Response Hilbert Transform for Edge Detection How to choose a edge detection filter? By Canny’s Theorem: Higher distinction Noise immunity 2019/1/3

Short Response Hilbert Transform for Edge Detection To have both the advantage, we have to satisfy: (Constraint 1) A1 < T < A2, (Constraint 2) (Constraint 3) if |x2| > |x1|  |x0|, (Constraint 4) h(x) = h(x) 2019/1/3

Short Response Hilbert Transform for Edge Detection The other alternative ways to define SRHLT: 2019/1/3

Improved Harri’s Algorithm For Corner And Edge Detections Comparision with SRHLT: More effective to defect corner More effective to defect edge Worse noise immunity 2019/1/3

Improved Harri’s Algorithm For Corner And Edge Detections Instead of we use: 2019/1/3

Improved Harri’s Algorithm For Corner And Edge Detections Basis: Harris’ Algorithm: x2, y2, xy Improved Agorithm:x2, y2, xy, x, y, 1 More types of edge detection 2019/1/3

Improved Harri’s Algorithm For Corner And Edge Detections (step 1) Find the orthonormal polynomial set that isorthogonal respect to a weighting function w[m, n]. (step 2) We do the inner product for L1[m, n, x, y] and k[x, y], where k = 1, 2, 3, 4, 5,6: 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections (step 3) Then we express the variation around [m, n] by: 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections (step 4) The principal axes are the eigen vector of 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections (step 5) we can observe the variation along the two principal axes. 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections Fig.17.Variations of gray levels along four principal directions for the pixel at a corner, on an edge, at a peak, and on a ridge[1] 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections Table 1 Case table.[1] 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections (step 6) Choose the best one 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections Using the proposed algorithm to do corner detection Using the proposed algorithm to do edge detection Using Harris’ algorithm to do corner detection Fig.18. Compare Harri’s Algorithm with the proposed algorithm.[1] 2019/1/3

Improved Harri’s Algorithm For Corner And Egde Detections Fig.19. (a) Image consists of three dots (upper-left), a valley (upper-right), a ridge (lower-left), and a noise-interfered region (lower-right). (b) (c) The results of corner detections.[1] 2019/1/3

Conlusion Fist-Order Derivative Edge Detection: the simplest method Second-Order Derivative Edge Detection: sensitive to noise Hilbert Transform for Edge Detection good for ramp edge, better noise immunity, but bad for accurate detection 2019/1/3

Conlusion Short Response Hilbert Transform for Edge Detection (SRHLT): have both advantage of HLT and differentiation Improved Harri’s Algorithm For Corner And Egde Detections: detect more types (34=81) 2019/1/3

Conclusion Application Image segmentation Data compression 2019/1/3

Reference [1] Soo-Chang Pei, Jian-Jiun Ding, ” Imporved Harris’ Algotithm ForConer And Edge Detections”, vol 1, 2005. [2] William K. Pratt , “Digital Image Processing_William K. Pratt 3rd ”,chapter 15. [3] Jiun-De Huang, “Image Compression by Segmentation and Boundary Description”, chapter 2. [4] J. Canny, “A Computational Approach to Edge Detection,” IEEE Trans. Pattern Analysis and Machine Intelligence, PAMI-8, 6,November 1986, 679–698. 2019/1/3

END 2019/1/3