Hello Imagej users,
I am just starting in image processing field. I recorded a process with a camera. The process is characterized by the appearance of a dark area in a very bright one. The position , the size and the shape of the dark area is variable. My aim is to identify a set of pictures I can use to describe and characterize the process by evaluating the number of pictures in the sequence that are looking similar to each type of picture of the set. I have already identified the set of picture. As criteria, I took the position and the size of the dark area. to obtain the position, used the following funtion: -smooth -find edges .binary-theshold invert -find maxima within a certain noise tolerance. Now I would like to implement a macro to classfy my sequence according to the similarity with each pictures of the
set. One example of conclusion I am looking for is for example" in a sequence of 20 pictures, and a set of 3 pictures taken as samples, 12 are similar to sample1, 3 to sample2 and 5 to sample 3" My question is doea anyone help me to define which criteria I should used to loop inside the sequence for a comparison? -I tried to plot the profile for each sample and compare it with the profile of each picture of the sample but the samples profiles are too similar so I can used the profile as criteria. .- the second Idea was to find the minima/maxima in each picture and compare the position with the position in each sample this does not work because the aim is to identify an area not a point. -third Idea, to track the shape of the dark area for each picture and compare with the samples, but the dark area shape and position are not so
stable and I dont know how to define a tolerance for comparison. can anyone offer me a suitable solution, any idea is welcomed thank in advance Paul Rostand when I tried the command
File<<open....
I got the error message plugin reader not found
after looking in the forum I knew about the plugin jimi reader and jimi writer. after dowloading and installing the plugins when I tried the command
run ("Jimi Reader", "open.....
I got the error message
<<Java.lang illegal state operation
can someone helps??
thanks in advance
regards
Paul Rostand
|
Hi Paul,
"looking similar" means what? Can you give an example? Would it be a solution to calculate the correlation of each pair of dark area objects after centering and align the objects according to their major axis. Regards, Peter Am 02.05.2011 21:52, schrieb paul_rostand: > Hello Imagej users, > > I am just starting in image processing field. I recorded a process with a > camera. The process is characterized by the appearance of a dark area in a very > bright one. The position , the size and the shape of the dark area is variable. > My aim is to identify a set of pictures I can use to describe and characterize > the process by evaluating the number of pictures in the sequence that are > looking similar to each type of picture of the set. > I have already identified the set of picture. As criteria, I took the position > and the size of the dark area. > to obtain the position, used the following funtion: > -smooth > -find edges > .binary-theshold > invert > -find maxima within a certain noise tolerance. > > Now I would like to implement a macro to classfy my sequence according to the > similarity with each pictures of the set. > > One example of conclusion I am looking for is for example" in a sequence of 20 > pictures, and a set of 3 pictures taken as samples, 12 are similar to sample1, 3 > to sample2 and 5 to sample 3" > > > My question is doea anyone help me to define which criteria I should used to > loop inside the sequence for a comparison? > > -I tried to plot the profile for each sample and compare it with the profile of > each picture of the sample but the samples profiles are too similar so I can > used the profile as criteria. > .- the second Idea was to find the minima/maxima in each picture and compare the > position with the position in each sample this does not work because the aim is > to identify an area not a point. > -third Idea, to track the shape of the dark area for each picture and compare > with the samples, but the dark area shape and position are not so stable and I > dont know how to define a tolerance for comparison. > > can anyone offer me a suitable solution, any idea is welcomed > > thank in advance > > Paul Rostand > > > > > when I tried the command > File<<open.... > > I got the error message plugin reader not found > after looking in the forum I knew about the plugin jimi reader and jimi writer. > after dowloading and installing the plugins when I tried the command > > > run ("Jimi Reader", "open..... > > I got the error message > > <<Java.lang illegal state operation > > can someone helps?? > thanks in advance > regards > > Paul Rostand > > -- > View this message in context: http://imagej.588099.n2.nabble.com/pattern-identification-and-classification-of-an-image-sequence-tp6325099p6325099.html > Sent from the ImageJ mailing list archive at Nabble.com. > |
Free forum by Nabble | Edit this page |