This movie shows a visualization of the visual buffer used in my visual SLAM algorithm. Two images are shown at the top. The left image shows all found SIFT keypoints, the right image shows only the SIFT keypoints that are left after pruning in the visual buffer.
At the bottom five smaller images are visible. These are the images currently in the buffer. The most left image is the same as the two larger images above (this is the oldest image in the buffer). The SIFT keypoints in this most left image are matched to the keypoints found in the other four images in the buffer. Only the keypoints are shown, that do match with at least three keypoints from the other four images.
Groningen!
v01d4d3pt 11 months ago