how do I solve matlab feature mismatching problem?

I am tring to match points on an object with two different view of it.Firstly I used minimum eigenvalu to find points and the results are below;

When I try to match two image the results are mismatch.

I couldn't find the reason.Could you help me to solve this problem?

My code;

I1 = rgb2gray(imread('IMG_1046.jpg')); C1 = corner(I1,'MinimumEigenvalue',800,'FilterCoefficients',fspecial('gaussian',[31 1],0.000000000000000009));

I2 = rgb2gray(imread('IMG_1048.jpg')); C2 =corner(I2,'MinimumEigenvalue',800,'FilterCoefficients',fspecial('gaussian',[51 1],0.000000000000009)); % imshow(I1); % hold on % plot(C1(:,1), C1(:,2), 'r.') imshow(I2); hold on plot(C2(:,1), C2(:,2), 'r.') [features1, valid_points1] = extractFeatures(I1, C1);

[features2, valid_points2] = extractFeatures(I2, C2);

indexPairs = matchFeatures(features1, features2); matched_points2 = valid_points2(indexPairs(:, 2), :); matched_points1 = valid_points1(indexPairs(:, 1), :); figure; showMatchedFeatures(I1, I2,matched_points1, matched_points2,'montage');

Answer 1

[Anand]

Take a look at the examples on this page. Part of the workflow described is feature matching:

http://www.mathworks.com/help/vision/gs/object-detection-and-tracking.html#btt5qyu

Also, here's a ready reference of the kinds of features detectors and descriptors available:

http://www.mathworks.com/help/vision/feature-detection-extraction-and-matching.html

Answer 2

[Dima Lisin]

Try using detectSURFFeatures() instead of corner().