http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703
MATLAB Central Newsreader  Hu monents
Feed for thread: Hu monents
enus
©19942014 by MathWorks, Inc.
webmaster@mathworks.com
MATLAB Central Newsreader
http://blogs.law.harvard.edu/tech/rss
60
MathWorks
http://www.mathworks.com/images/membrane_icon.gif

Wed, 20 Jul 2011 20:00:10 +0000
Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846105
joveria
Hi,<br>
I am using <br>
<br>
mji=sum(sum array(x,y).(xx')^j.(yy')^i)<br>
<br>
formula to compute hu moments. I want to know that I am doing this right or is there any other formula to compute hu moments?<br>
Or is this the right formula for calculating hu moments?

Thu, 21 Jul 2011 02:53:09 +0000
Re: Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846146
Roger Stafford
"joveria" wrote in message <j07c4a$1bo$1@newscl01ah.mathworks.com>...<br>
> Hi,<br>
> I am using <br>
> <br>
> mji=sum(sum array(x,y).(xx')^j.(yy')^i)<br>
> <br>
> formula to compute hu moments. I want to know that I am doing this right or is there any other formula to compute hu moments?<br>
> Or is this the right formula for calculating hu moments? <br>
           <br>
See <a href="http://en.wikipedia.org/wiki/Image_moment">http://en.wikipedia.org/wiki/Image_moment</a><br>
<br>
Roger Stafford

Thu, 21 Jul 2011 10:43:08 +0000
Re: Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846180
joveria
<br>
when hu moments give values in exponentials then what does it mean? is it giving right or wrong answer. plz guide me in this<br>
thanks in advance

Thu, 21 Jul 2011 16:58:09 +0000
Re: Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846237
Roger Stafford
"joveria" wrote in message <j08vrs$899$1@newscl01ah.mathworks.com>...<br>
> <br>
> when hu moments give values in exponentials then what does it mean? is it giving right or wrong answer. plz guide me in this<br>
> thanks in advance<br>
         <br>
What you have shown in your first posting is the formula for an image's central moments. If you consult the Wikipedia referenc I gave you, these can be used to compute the eta scale invariant moments, and these in turn can be used to compute the eight different Hu moments in accordance with the formulas given there.<br>
<br>
I don't understand your question where you wrote, "hu moments give values in exponentials". The only exponentials involved are the "orders" associated with the various moments. What exactly are you asking?<br>
<br>
Roger Stafford

Thu, 21 Jul 2011 17:26:09 +0000
Re: Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846244
joveria
<br>
i have computed moment invariants by<br>
m02=sum(sum(double(im).*(rowcentroid(1))^0.*(colcentroid(2))^2)<br>
<br>
then normalized them by using this <br>
<br>
n02=m02/area^gamma<br>
gamma=2+(0+2)/2<br>
<br>
n i got value like this: 3.456e4<br>
<br>
my question is: does normalizing give value in exponentials as i got to know that normalization should not be in exponentials

Thu, 21 Jul 2011 18:04:08 +0000
Re: Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846252
Roger Stafford
"joveria" wrote in message <j09nfh$jcf$1@newscl01ah.mathworks.com>...<br>
> <br>
> i have computed moment invariants by<br>
> m02=sum(sum(double(im).*(rowcentroid(1))^0.*(colcentroid(2))^2)<br>
> <br>
> then normalized them by using this <br>
> <br>
> n02=m02/area^gamma<br>
> gamma=2+(0+2)/2<br>
> <br>
> n i got value like this: 3.456e4<br>
> <br>
> my question is: does normalizing give value in exponentials as i got to know that normalization should not be in exponentials <br>
        <br>
Your question is still not clear to me, Joveria. You ask, "does normalizing give value in exponentials" and "normalization should not be in exponentials". I don't know what you mean by that. What "exponentials" are you referring to? In what way is that related to the 3.456e4 answer you received? There seems to be a language barrier here. Can you give some concrete set of numbers in terms of which to pose that question?<br>
<br>
A second point: In your code<br>
<br>
n02=m02/area^gamma<br>
gamma=2+(0+2)/2<br>
<br>
I assume that 'area' is just m00. If so, it looks as if the gamma should be:<br>
<br>
gamma = 1+(0+2)/2<br>
<br>
rather than what you have. See the formula for eta_i_j in the Wikipedia reference.<br>
<br>
Roger Stafford

Fri, 22 Jul 2011 05:06:10 +0000
Re: Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846305
sarah
Sorry Roger, i am not able to explain you my problem. By exponential i mean in the power of 10. I am getting values like this:<br>
I1=0.2812,0.0073,24.8925,15.4261,837.1579<br>
I2=0.0761,0.0280,845.1402,355.8602,7.0071e+05<br>
I3=0.007,3.2219e04,506.2985,94.0678,3.7975e+07<br>
I4=0.0043,1.7355e04,360.7009,57.3486,3.7911e+07<br>
I5=2.5865e05,7.8553e09,6.1117e+05,1.4601e+15<br>
I6=6.7652e+09,8.3733e+08,5.4083e+18,9.7474e+17,3.1191e+20<br>
<br>
these are values of differents objects in one image: hope you now understand

Sat, 23 Jul 2011 08:32:11 +0000
Re: Hu monents
http://www.mathworks.com/matlabcentral/newsreader/view_thread/310703#846442
joveria
Sorry Roger, i am not able to explain you my problem. By exponential i mean in the power of 10. I am getting values of hu moments like this:<br>
I1=0.2812,0.0073,24.8925,15.4261,837.1579<br>
I2=0.0761,0.0280,845.1402,355.8602,7.0071e+05<br>
I3=0.007,3.2219e04,506.2985,94.0678,3.7975e+07<br>
I4=0.0043,1.7355e04,360.7009,57.3486,3.7911e+07<br>
I5=2.5865e05,7.8553e09,6.1117e+05,1.4601e+15<br>
I6=6.7652e+09,8.3733e+08,5.4083e+18,9.7474e+17,3.1191e+20<br>
<br>
these are values of differents objects in one image: hope you now understand now<br>
now my q is: are these values are correct? means can i get these types of values?