IBMR Assignment 1

Post on 06-Feb-2016

33 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

DESCRIPTION

IBMR Assignment 1. Stitching Photo Mosaics. What is Photo Mosaic?. Stitching photos to construct a wild-view scene. Part1: CORNER DETECTION Part2: PERSPECTIVE MAPPING and MOSAICING Handout after Part2 Finished. CORNER DETECTION. Requirement. - PowerPoint PPT Presentation

Transcript

IBMR Assignment 1Stitching Photo Mosaics

What is Photo Mosaic?• Stitching photos to construct a wild-view scene.• Part1: CORNER DETECTION• Part2: PERSPECTIVE MAPPING and

MOSAICING• Handout after Part2 Finished

CORNER DETECTION

Requirement• Read an image, detect its all corner by HARRIS

CORNER DETECTION• (bonus) Invariant or robust features on scale,

orientation, illumination

Support• A GDI+ Image Loader• You may use any other library(openCV, matlab,

etc) except existing corner detection function.

About GDI+• Get color from Pixel(X,Y) in gbmpPicL• Color c;• gbmpPicL->getPixel(X, Y, &c);• int red = c.GetR();

• Set color at Pixel(X,Y) in gbmpPicL• gbmpPicL->setPixel(X, Y, Color(0,255,0)); //draw green at

(X,Y)

About Harris Detection• Preprocessing: Gaussian Filter / Gray level• Consider the matrix for a small square around (x,y)• Compute x and y derivatives : and • Compute • Find λ1 and λ2 by evaluation of eigenvalues or SVD

or• Compute the response of the detector at each pixel: Positive: corner Negative: edge Small: flat

k~0.04 to 0.06 + and

Next Step• Show the response

cited from http://www.ikaros-project.org/articles/2004/monoculardepth1/

• Non-Maxima suppression

cited from http://ssip2003.info.uvt.ro/lectures/chetverikov/edge_detection.pdf

Result – Try your own photo!•

top related