yoonsejong
commented
3 years ago Hi,
1.Is there a step missing in the readme where one is excepted to download the distfcm function from this link?
No. It's mentioned in Step 7. To calculate the Wang's fuzzy C-Mean distance, you need to first find centroids from training data. My codebase has train/predict code to train the centroid. The distfcm function you mentioned merely calculates the distance after you have centroid calculated.
- I'm not seeing where colorcalc.mex is used in the code base, and I had troubles generating the mex file.
It can be downloaded (and compiled) from https://www.mathworks.com/matlabcentral/fileexchange/28790-colorspace-transformations as I explained in Step 4. I just downloaded and compiled again on my computer (Windows), and it correctly creates the needed mex file. It's just an x64 version. [image: image.png]
- Assuming Yes to the first question and omitting the colorcalc.mex* file, without which the code can still ran, I ran collect_MH_features.m on some random image, and ran into the following error:
It is possible that some feature values end up being zero. Regarding the initial error about index, I suggest you try square images instead of rectangular ones (not required, but often solves minor technical issues).
Thanks.
On Wed, Jun 30, 2021 at 1:36 AM WenXinDong2018 @.***> wrote:
Hi,
After downloading all the necessary files, I ran into some errors.
1.Is there a step missing in the readme where one is excepted to download the distfcm function from this link https://www.mathworks.com/matlabcentral/mlc-downloads/downloads/submissions/48686/versions/1/previews/FuzzyClusterToolBox/FCM_Matlab/distfcm.m/index.html ?
- I'm not seeing where colorcalc.mex is used in the code base, and I had troubles generating the mex file.
- Assuming Yes to the first question and omitting the colorcalc.mex* file, without which the code can still ran, I ran collect_MH_features.m on some random image, and ran into the following error:
Index in position 1 is invalid. Array indices must be positive integers or logical values.
Error in itten (line 89) xy = [xy(end,:); xy; xy(1,:)];
Error in collect_MH_Features (line 144) [fullfeatures(22:41), fullfeatures(98)] = itten(img);
To temporarily bypass this error, I replaced feature 6 and feature 16 in the Itten file with 0, and here is a feature vector I got for a random image. As can be seen, feature #28 and #38 are 0s (corresponding to Itten feature #6 and #16).
0.222700081299947 0.718211096870800 0.544559674726840 -0.0890253912499798 0.617104459637791 0.530720948102808 0.960960938176237 -0.104241655748984 0.939209460786395 60.8022981261836 0.0395904961388000 0.350459764083375 0.0344600633341871 0.0537428277679468 0.0363860311588302 0.0370301977439113 0.0754686947888207 0.138995025930227 0.0437309355431238 0.154810592449058 0.0353253710617197 0.150057400580607 0.151313467836416 0.0452853736462183 -26.0821144430514 1 0 0 949.759164550661 NaN NaN 0.150057400580607 0.151313467836416 1.99998462944387 3.12465553658501 1 0 0 17193.0690275822 0 0 0.000865069306133945 0.000241414690083892 0.000160943126722594 0.000422475707646810 0.00237391111915827 0.00191119962983081 0.00388275293218259 0.00331945198865351 0 0 0.000100589454201622 0.000181061017562919 0.00456676122075362 0.00460699700243426 0.00374192769630032 0.00171002072142757 0.000345729989059788 9.34445312749523e-07 3.57100412435505 0.00710161546663448 0.0279236324863701 0.251393163940692 0.270525278129841 0.443056309976462 0.00728267648419740 0.00589454201621502 0.114685789116378 0.00266540212844503 0.277564147124758 0.918507854428645 0.355986625204815 1 0.105633981663518 0.000433019566146501 -3.08162637729203e-05 0.610433538510885 0.000927437428985417 0.00258242559756492 -0.135976513751465 0.000278325046126904 -0.000147918503924695 0.106036184965892 0.613943401537435 -0.135846107209263 0.212765957446809 0.117001051694847 0.332618720168271 0.392571061002011 0.979433635207481 0.943362464563498 0.988114290746447 0.251786206766153 0.175454114227190 0.996084796267832 0.944277027209234 0.910080764770919 35 3.21227828412621 -0.901779026217229 9.38510167745250 -20.5714285714286 0.385714285714286 21.9186196887142 -44 0.255555555555556 41.5148542550203 2036.31890229414 1.52629233511586 7.06526087344029 0 0 0.0454463154082926 1
I would greatly appreciate your help.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yoonsejong/mh_huevent2014/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEGOUX5LO5PF6RBMJXDQL4DTVKUOXANCNFSM47RSTLBQ .
WenXinDong2018
Hi,
After downloading all the necessary files, I ran into some errors.
1.Is there a step missing in the readme where one is excepted to download the
distfcmfunction from this link?colorcalc.mexis used in the code base, and I had troubles generating the mex file.colorcalc.mex*file, without which the code can still ran, I ran collect_MH_features.m on some random image, and ran into the following error:To temporarily bypass this error, I replaced feature 6 and feature 16 in the Itten file with 0, and here is a feature vector I got for a random image. As can be seen, feature #28 and #38 are 0s (corresponding to Itten feature #6 and #16).
I would greatly appreciate your help.