analyzing a line, please

[Jimena Berni]

Hello,

sorry the question is a little bit basic but I haven't found a way of  
doing it, so i would like to ask for help.

I am trying to analyze the track of two different animals (image  
attached). One of them, the treatment is going kind of in circles  
while the other perform straight lines and then turns.

I thought i could draw a segmented line over the track and get the  
value of angle between two adjacent points. I don't find a way of  
doing this.
I thought that another alternative would be to get the XY information  
of the line. Is there a way of doing that?

Any other idea will be more then welcomed.

Many thanks!


Jimena


------------
Dr. Jimena Berni
Henslow Research Fellow
Downing College
Department of Zoology
University of Cambridge
Downing Street
CB2 3EJ
Cambridge, UK

Tel.: (++44) 1223 336635
Fax.: (++44) 1223 336676
[hidden email]

[dscho]

Hi Jimena,

On Fri, 8 Jul 2011, Jimena Berni wrote:

> I am trying to analyze the track of two different animals (image
> attached). One of them, the treatment is going kind of in circles while
> the other perform straight lines and then turns.
>
> I thought i could draw a segmented line over the track and get the value
> of angle between two adjacent points. I don't find a way of doing this.
> I thought that another alternative would be to get the XY information of
> the line. Is there a way of doing that?

Of course that is one way to do it, although the analysis would
definitely require programming a macro.

But if all you seek is to have a robust discrimination between treated vs
untreated animals, I think it might be better to measure enclosed area vs
average intensity. At least to me it looks like the "cuddled" paths cover
a smaller area, so the average brightness in that area is much higher than
in the other case.

Ciao,
Johannes

[chai_san]

Hi Jimena,
Alternatively a skeletonization routine will give you path lengths and all sorts of neat parameters- directed or not, random or not etc. I remember seeing some plugins for this.
Cheers
Chaitanya


On 08-Jul-2011, at 10:54 PM, Johannes Schindelin wrote:

> Hi Jimena,
>
> On Fri, 8 Jul 2011, Jimena Berni wrote:
>
>> I am trying to analyze the track of two different animals (image
>> attached). One of them, the treatment is going kind of in circles while
>> the other perform straight lines and then turns.
>>
>> I thought i could draw a segmented line over the track and get the value
>> of angle between two adjacent points. I don't find a way of doing this.
>> I thought that another alternative would be to get the XY information of
>> the line. Is there a way of doing that?
>
> Of course that is one way to do it, although the analysis would
> definitely require programming a macro.
>
> But if all you seek is to have a robust discrimination between treated vs
> untreated animals, I think it might be better to measure enclosed area vs
> average intensity. At least to me it looks like the "cuddled" paths cover
> a smaller area, so the average brightness in that area is much higher than
> in the other case.
>
> Ciao,
> Johannes

------------------------------------------
Chaitanya Athale Dr. rer. nat.
Asst. Prof.
Biology
IISER Pune
www.iiserpune.ac.in/~cathale
------------------------------------------

[iqm mug]

I would suggest a fractal analysis. The fractal dimension of the treated
images may have higher values.

Best wishes,
Helmut

2011/7/8 Johannes Schindelin <[hidden email]>

> Hi Jimena,
>
> On Fri, 8 Jul 2011, Jimena Berni wrote:
>
> > I am trying to analyze the track of two different animals (image
> > attached). One of them, the treatment is going kind of in circles while
> > the other perform straight lines and then turns.
> >
> > I thought i could draw a segmented line over the track and get the value
> > of angle between two adjacent points. I don't find a way of doing this.
> > I thought that another alternative would be to get the XY information of
> > the line. Is there a way of doing that?
>
> Of course that is one way to do it, although the analysis would
> definitely require programming a macro.
>
> But if all you seek is to have a robust discrimination between treated vs
> untreated animals, I think it might be better to measure enclosed area vs
> average intensity. At least to me it looks like the "cuddled" paths cover
> a smaller area, so the average brightness in that area is much higher than
> in the other case.
>
> Ciao,
> Johannes
>

[dscho]

Hi,

On Sat, 9 Jul 2011, Chaitanya Athale wrote:

> Alternatively a skeletonization routine will give you path lengths and
> all sorts of neat parameters- directed or not, random or not etc. I
> remember seeing some plugins for this.

As far as I saw, the paths were overlapping. If you have a skeletonization
routine that handles such paths correctly, I would be _very_ interested!

Thanks,
Johannes

> On 08-Jul-2011, at 10:54 PM, Johannes Schindelin wrote:
>
> > Hi Jimena,
> >
> > On Fri, 8 Jul 2011, Jimena Berni wrote:
> >
> >> I am trying to analyze the track of two different animals (image
> >> attached). One of them, the treatment is going kind of in circles while
> >> the other perform straight lines and then turns.
> >>
> >> I thought i could draw a segmented line over the track and get the value
> >> of angle between two adjacent points. I don't find a way of doing this.
> >> I thought that another alternative would be to get the XY information of
> >> the line. Is there a way of doing that?
> >
> > Of course that is one way to do it, although the analysis would
> > definitely require programming a macro.
> >
> > But if all you seek is to have a robust discrimination between treated vs
> > untreated animals, I think it might be better to measure enclosed area vs
> > average intensity. At least to me it looks like the "cuddled" paths cover
> > a smaller area, so the average brightness in that area is much higher than
> > in the other case.
> >
> > Ciao,
> > Johannes
>
> ------------------------------------------
> Chaitanya Athale Dr. rer. nat.
> Asst. Prof.
> Biology
> IISER Pune
> www.iiserpune.ac.in/~cathale
> ------------------------------------------
>

[Jimena Berni]

Thank you all for your answers and suggestions!

I'll try a fractal analysis.

Jimena

------------
Dr. Jimena Berni
Henslow Research Fellow
Downing College
Department of Zoology
University of Cambridge
Downing Street
CB2 3EJ
Cambridge, UK

Tel.: (++44) 1223 336635
Fax.: (++44) 1223 336676
[hidden email]








On 9 Jul 2011, at 13:01, iqm mug wrote:

> I would suggest a fractal analysis. The fractal dimension of the  
> treated
> images may have higher values.
>
> Best wishes,
> Helmut
>
> 2011/7/8 Johannes Schindelin <[hidden email]>
>
>> Hi Jimena,
>>
>> On Fri, 8 Jul 2011, Jimena Berni wrote:
>>
>>> I am trying to analyze the track of two different animals (image
>>> attached). One of them, the treatment is going kind of in circles  
>>> while
>>> the other perform straight lines and then turns.
>>>
>>> I thought i could draw a segmented line over the track and get the  
>>> value
>>> of angle between two adjacent points. I don't find a way of doing  
>>> this.
>>> I thought that another alternative would be to get the XY  
>>> information of
>>> the line. Is there a way of doing that?
>>
>> Of course that is one way to do it, although the analysis would
>> definitely require programming a macro.
>>
>> But if all you seek is to have a robust discrimination between  
>> treated vs
>> untreated animals, I think it might be better to measure enclosed  
>> area vs
>> average intensity. At least to me it looks like the "cuddled" paths  
>> cover
>> a smaller area, so the average brightness in that area is much  
>> higher than
>> in the other case.
>>
>> Ciao,
>> Johannes
>>