Please enable JavaScript.
Coggle requires JavaScript to display documents.
5700_3 (Linear Filters (Shift invariant linear systems (Superposition,…
5700_3
Linear Filters
kernel
convolution
blurring
sharpening
Smoothing
Average filter (box filter)
Gaussian
Shift invariant linear systems
Superposition
Scaling
Shift invariance
Differentiation and convolution
Spatial frequency and Fourier Transform
Sampling and aliasing
Smoothing and resampling
Edge and Corner Detection
Smoothing and Differentiation
Scale
Increased smoothing:
• Eliminates noise edges.
• Makes edges smoother and thicker.
• Removes fine detail.
Derivative of Gaussian
Criteria for optimal edge detection:
Good detection
Good localization
Single response constraint
Canny Edge Detection
Smooth image w/ Gaussian filter
Apply derivative of Gaussian
Find magnitude and orientation of gradient
‘Non-maximum suppression’
‘Hysteresis’: Linking and thresholding
Use khigh to find strong edges to start edge chain
Use klow to find weak edges which continue edge chain
Morphological Image Processing
Basic Concepts from Set Theory
Erosion and Dilation
膨胀是一种非线性操作；
卷积是一种线性操作
桥接：形态学可以直接在一幅二值图中得到结果；低通滤波方法从二值图开始，生成灰度图，再通过阈值将灰度图转化为二值图
Opening and Closing
Hit or miss
Gray-Scale Morphology
Erosion and Dilation
Opening and Closing
Some Basic Gray-Scale Morphological Algorithm
Morphological smoothing
Morphological gradient
Top-hat and bottom-hat transformations
Gray-Scale Morphological Reconstruction
Some Basic Morphological Algorithms in binary images
Boundary Extraction
Hole Filling
Skeletons
Morphological Reconstruction
Image Segmentation
Defi nition
Basic Idea
Detection of Discontinuity
Point detection
Laplacian
Line detection
Laplacian
Edge detection
Gradient : fi rst order derivative
Laplacian: Second order derivative
Laplacian of Gaussian
Gaussian Derivative
Canny edge detector
Edge Linking
Local processing
Global processing (based on Hough Transform)