Visualization to help choose sigmas for meanShiftIm. Displays the original image I, and prompts user to select a point on the image. For given point, calculates the distance (both spatial and range) to every other point in the image. It shows the results in a number of panes, which include 1) the original image I, 2) Srange - similarity based on range only, 3) Seuc - similarity based on Euclidean distance only, and 4) overall similarity. Finally, in each image the green dot (possibly occluded) shows the original point, and the blue dot shows the new mean of the window after 1 step of meanShift. USAGE meanShiftImExplore( I, X, sigSpt, sigRng, [show] ) INPUTS I - MxN image for display X - MxNxP data array, P may be 1 (X may be same as I) sigSpt - integer specifying spatial standard deviation sigRng - value specifying the standard deviation of the range data show - [1] will display results in figure(show) OUTPUTS EXAMPLE I=double(imread('cameraman.tif'))/255; meanShiftImExplore( I, I, 5, .2, 1 ); See also MEANSHIFTIM Piotr's Computer Vision Matlab Toolbox Version 2.0 Copyright 2014 Piotr Dollar. [pdollar-at-gmail.com] Licensed under the Simplified BSD License [see external/bsd.txt]

Generated by