Denisty Estimation + Mean shift, Markov Random Fields,Feature and corner detection Flashcards
Limitations and beyond Kmeans
Limitations
1-Needs k to be provided
2-Non Optimal
3-Forces symetry
Beyond:
1-Create Density Function
2-Look For modes
Creating density function for segmentatiton
Starting point: set of samples
Approac: Density estimation
Density function:computionally complex
Mean Shift
-Find stationary peaks without computing distribution function.
-models distribution using continuos nonparametric model
Avoid density function? Estimate gradient instead.
Meanshift steepest ascend method
Meanshfit vector same direction as gradient.
Procedure:
-Compute mean shift vector
-move kernel window
-stop if gradient close to 0
-prube mode by perturbation
Meanshift properties and optimization
Properties:
1-Automatic convergence speed
2-Steps are smaller towards mode.
3-Convergence depends on step size
Optimization
1-Divide space into windows and run parallelly
Pros
1.Doesnt assume cluster spherical
2.1 parameter window size
3- Find a variable nr of modes
4-robust to outliers
Cons
1-depends on window size -> hard to find the best, if bad modes are merged
2-computionally expensive
3-doesnt scale well with dimension of feature space
Mean shift Clustering
Clustering criterion
1. Define attraction basin as the region for which trajectories lead to same mode.
- All data points in same basin = cluster
To perserve discontinuities we have a joint domain spatial+color
Segmentation and optimization
1 additional approach -> segmentation is a per-pixel labelling task.
Labelling::
-Define a set of labels
-Define and energy function
-Assign a label to minimize the energy
Energy function:
1-Data term
Depends on specific image features at pixel location
Evaluates how good label fits pixel
Defines match criterion based on image feature vector.
2.Smoothing term
Depends on labels of neighbors.
Normally 4 neighbors considered
Favors same label at adjacen pixels.
Balance:
1-To detect strong boundaries: dont penalize with smoothing term
2-Weak boundaries like noise shouldnt affect labelling so smoothing term shouldnt be too weak.
Markov Random FIelds
The energy function evaluated over the whole
image is defined as:
πΈ π = ΰ·
πβΞ©
πΈdata π, ππ + ΰ·
πβπ΄(π)
πΈsmooth(ππ, ππ)
* This model is called Markov Random Field (MRF)
* The model considers local interactions between
adjacent pixels
-Interactions are modeled considering a graph
Goal: select the labelling that minimizes the
energy function
Optimizing Random Markov Fields using Belief Propagation
MRF-> minimization problem
Belief Propagation:algorithm to solve it
Based on message passing between neighboring pixels
Message has info about labels
Run several iterations
A message is a function mapping
β The ordered discrete set of labels β πΏ
β Into an array of length π having, for
each position, the message value
* Non-negative real values
β Message value for label π sent from π
to π at time π‘
β Function of the label π β πΏ
The reply of node π depends on
β The data term at π for label β (minimized)
β The smoothing term if the labels π and β are different
β The information brought by the neighbors of π
(excluding π) in the previous iteration (time π‘ β 1)
β Node π is excluded from the computation
* It cannot influcence the opinion of π about π itself
Features
Feature is a meaningful part of image.
Output: a set of points + description of the region (AKA
signature, fingerprint, β¦)
The ideal keypoints shall be
β Stable and repeatable
β Invariant to transformations (e.g., rotations)
β Insensitive to illumination changes
β Accurate
The ideal descriptor shall be
β Robust to occlusion and clutter
β Robust to noise, blur, compression, discretization
β Discriminative
β Stable over changes in viewing angle and illumination
β Computationally efficient (many features per image)
Harris Corners to detect salient points(corners)
Idea: Consider image and shifted version
if uniform region:two similar patches.
if salient point; different patches.
Corner = large difference.
Harris.
- Conisder a patch and a displacement.
- Similarity is measured with autocorrelation.
Autocorrelation matrix properties.
-real
-symetric
-orthogonal eigenvectors
-points to max data spread eigenvectors
Studying the eigenvalues we get information
about the type of patch
β If both eigenvalues are small: uniform region
β Only one large eigenvalue: edge
β Two large eigenvalues: corner
The Harris corner detector is
β Invariant to brightness offset:
β Invariant to shift and rotations
β Not invariant to scaling
Beyond Harris Corners
Susan and Usan
-Analyze circular window around point
-no derivatives
-edge and corner detector
-robust to noise
USAN: Comparte between nucleus and pixels. The portion with intensity difference from nucleus given a threshold.
Susan i smalles usan.
Harris focuses on specific point.
OTHERS on BLOBS(diferent properties than surrounding).
MSER
Connected areas with uniform intensity surrounded by contrasting background (blobs)
MSER DETECTOR IS SAME AS BLOB DETECTOR.
ALG:
Apply thresholds
compute connected binary regions
compute statistics
analyze how presisten is blob.
Scale Space Concept
Goal: analyze image at multi scales
Tool: scale space 0> original signal to family of derived signals.
-fine details are gradually suppressed.
Requirements:
Linearity
Spatial shift invariance
no new local extrema created
scale invariance
Gaussian smoothing does it
Edge detection in scale space
Strong edges found at all scales.
At fine scales,accurate but many.
At coarse scales, inacurate but strong.
Combine them.
SIFT Feature in detail
Keypoint detection
Descriptor calculation
Invariant to: translation, rotation, scale and other
transforms
Application
Matching and
registration
* Object recognition
* Query by example
-image content mapped into local fature cordinates
SIFT features:
β Local β robust to occlusions
β Distinctive β distinguish objects in large databases
β Dense β many features can be found even on
small objects
β Efficiency
Algorithm
* Scale-space extrema detection
* Keypoint localization
* Orientation measurement
* Descriptor calculation