Changes between Version 29 and Version 30 of GSoC/2016/Additional_segmentation_algorithms/weekly_report

Timestamp:

May 30, 2016, 7:23:36 PM (8 years ago)

Author:

hao2309

Comment:

--

GSoC/2016/Additional_segmentation_algorithms/weekly_report

@@ -37 +37 @@
 
 === What did you get done this week? ===
-* Further discussion about the algorithm mechanism (about the edge effect)
-1. Edge effect:the moving window have to be re-size when it near the edge or corner of the full image. Mentor has given the solution:
+* Further discussion about the algorithm mechanism
+1. Edge effect:the moving window have to be re-size when it near the edge or corner of the full image. Mentor has given the solution in pseudo-code:
 {{{
       # figure out moving window, clip to region if necessary
@@ -47 +47 @@
       if (mwrow2 > nrows)
         mwrow2 = nrows
-
       
       mwcol1 = col - (int)radius
@@ -56 +55 @@
 }}}
 2. Adaptive bandwidth:
+Earlier mean-shift algorithm uses fixed bandwidth. Fixed bandwidth could result in either over-segment or under-segment. In the literature of Deng et al., 2015  and [http://doi.org/10.1049/iet-ipr.2013.0195 Zhang et al., 2014], each paper proposed an adaptive bandwidth method. [http://doi.org/10.1049/iet-ipr.2014.0393 Zhou et al., 2015] proposed another method, which use smaller bandwidth at first to over-segment the image, then use image clustering, region-based mode merging again to refine the result.
 3. The convergence condition:
 * Got the access for GRASS-addons-svn and [https://trac.osgeo.org/grass/browser/sandbox/bo/i.segment.gsoc2016 sandbox]