Up: Component summary Component

MaskClusterDraw

Reads cluster id:s, images and masks to create a colored visualization of object clustering. The coloring matches Plot2D coloring, if colorFunction="rainbow(%d)" has been used.

Example:

Version 1.0
Bundle anima
Categories
Authors Ville Rantanen (ville.rantanen@helsinki.fi)
Issue tracker View/Report issues
Requires Matlab
Source files component.xml imageclusters.m execute.m
Usage Example with default values

Inputs

Name Type Mandatory Description
mask ImageList Mandatory A directory with BW mask image files.
clusters CSV Mandatory A CSV file with file name, object number and cluster id number.
images ImageList Optional A directory with image files to be overlayed on.
names ImageList Optional A directory with image files with names used in the cluster id list.

Outputs

Name Type Description
out ImageList A directory with image files.

Parameters

Name Type Default Description
alternateWidth boolean true Draw even clusters with thick line. Useful if there are many clusters. Does not apply if fill is enabled
clusterId string "clusterId" Column name of cluster ids
colorFunction string "hsv(u)" Function used to colorify the clusters. Variable u is the number of unique clusters. Must return ux3 RGB value matrix. See http://www.mathworks.se/help/matlab/ref/colormap.html
fileId string "File" Column name of file name
fill boolean false Fill objects instgead of showing just the perimeter
leaveTrail boolean false Assume data is time lapse, do not erase the old objects from previous frame.
objectId string "Object" Column name of object number
useColor boolean true Use colors for clusters (not much point if there are more than 1 clusters)

Test cases

Test case Parameters IN
mask
IN
clusters
IN
images
IN
names
OUT
out
case1 properties mask clusters (missing) (missing) out

clusterId=clusterId

case2 properties mask clusters images (missing) out

clusterId=clusterId


Generated 2018-12-17 07:42:16 by Anduril 2.0.0