Mean Shift Segmantation

Part 1: Download NYU Depth Dataset [1] (specifically, you need to download Labeled dataset [2]). Choose an input pair (an RGB image and a corresponding depth map) other than the input pair shown in Figures 1-6. Convert the image from RGB color space to Lab color space using rgb21ab function. Represent each pixel via its Lab color values. Construct a feature matrix X where rows correspond to pixels and columns correspond to features. Normalize the feature matrix X so that, each of its columns has zero mean and standard deviation one. Apply mean shift algorithm using different values of the window size (or bandwidth) parameter. You can use the mean shift algorithm implementation [3] (which can be downloaded from Moodle). This code will return the cluster or segment id of each pixel. Compute the mean RGB value for each segment. Visualize the segmentation result as an RGB image where each pixel is displayed with the mean RGB value of its segment id. Discuss the results in terms of how they change as the window size or bandwidth parameter changes