ece734 project-scale invariant feature transform algorithm

10
ECE734 Project- Scale Invariant Feature Transform Algorithm Jing Li

Upload: milek

Post on 09-Feb-2016

57 views

Category:

Documents


0 download

DESCRIPTION

ECE734 Project-Scale Invariant Feature Transform Algorithm. Jing Li. 1. Background. 2. Algorithm introdution. 3. Program implement. Outline. Background: Scale-Invariant Feature Transform (SIFT). Scale-Invariant Feature Transform: Apply in image recognition. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: ECE734 Project-Scale Invariant Feature Transform Algorithm

ECE734 Project-Scale Invariant Feature Transform AlgorithmJing Li

Page 2: ECE734 Project-Scale Invariant Feature Transform Algorithm

Outline

Background1

Algorithm introdution2

Program implement3

Page 3: ECE734 Project-Scale Invariant Feature Transform Algorithm

Background: Scale-Invariant Feature Transform (SIFT)

• Scale-Invariant Feature Transform: Apply in image recognition.

• Two steps for image recognition: 1. Feature extraction 2. Feature match• SIFT algorithm can extract distinctive features.• Features: invariant to image scale and rotation perform reliable matching between different views of object or

scene. • Random Sample Consensus (RANSAC): Match

features from object image to training images

Page 4: ECE734 Project-Scale Invariant Feature Transform Algorithm

•Scale-space extrema detection

•Orientation assignment

•Keypoint localization

•Keypoint descriptor

Page 5: ECE734 Project-Scale Invariant Feature Transform Algorithm

Scale-space extrema detection

• Define image scale space function L(x, y,σ) = G(x, y, σ) ∗ I(x, y)• Detect stable keypoint locations

used difference-of-Gaussian function

D(x,y,σ)= L(x, y,kσ) - L(x, y,σ) • D(x,y,σ) is the approximation of

Laplacian of Gaussian (LoG) operation.

• We use five steps to build the use Build difference-of-Gaussian Octaves D(x,y,σ)

use different to conversion to build each octave of scale space, based on the function:

Page 6: ECE734 Project-Scale Invariant Feature Transform Algorithm

Keypoint localization

• Maxima and minima of LoG produce the most stable image features compared to a range of other possible image functions

• Detected by comparing a pixel (marked with X) to its 26 neighbors in 3x3 regions at the current and adjacent scales

• Talyor expansion for D(x,y,σ) and set its derivative to 0

Page 7: ECE734 Project-Scale Invariant Feature Transform Algorithm

Orientation Assignment

• Assign orientation to each keypoints by local gradient direction

• Gradient magnitude, m(x, y),• Orientation θ(x,y)

• Divide image into equal regions (4,16…….)

• Divide each region to some small regions.

• Find the orientation of small region and shift to the location of keypoint of large region. Overlay these orientation and use it to express the feature of keypoints.

Keypoint Descriptor

Page 8: ECE734 Project-Scale Invariant Feature Transform Algorithm

Program Test

Pgm file: Keypoint descriptor by SIFT algorithm1021 keypoints found

An image of size 500 pixels square will typically giveover 1000 keypoints depending on image content

Page 9: ECE734 Project-Scale Invariant Feature Transform Algorithm

• SIFT extract features from training image and store in database. Set a sample image which include the same features with the training image. Compare the Euclidean distance of keypoint to test corrective of algorithm

• Same features typically have the minimum Euclidean distance.• Compare Euclidean distance of each keypoints.

Page 10: ECE734 Project-Scale Invariant Feature Transform Algorithm