Take the 2-minute tour ×
Biology Stack Exchange is a question and answer site for biology researchers, academics, and students. It's 100% free, no registration required.

I am trying to recognize a deer by its antlers or any other means. Elaborating: I was hoping to use their antlers to recognize them but I have heard that most deers shed their antlers every year so it would be difficult to recognize it from the last year's photo unless these antlers retain the same pattern every year.

If not the antlers, what other characteristics should I be looking for?

Is there any software that can help me in recognizing a deer?

enter image description here

share|improve this question
2  
i m not sure on what basis you want to identify a deer.. but you can use machine learning algorithms to find your deer (you have to train your classifier first) –  WYSIWYG Apr 15 '13 at 19:13
1  
I think the main question you have written does not quite fall within the scope of biology SE. However, your final question, "What would it take to recognize a deer by its photo?" could be answered in biological terms. Perhaps that might be the better title. –  dd3 Apr 15 '13 at 21:04
1  
Changed the title, provided more explanation and even though I am a software engineer, my question is now completely biological. Please up vote me now. If zoology experts cannot not help me then who can? –  RHT Apr 15 '13 at 21:51
1  
One part of this question remains unanswered- does the shape of a deer's antlers act as some kind of identifier & are they consistent between years? I guess to answer this one would have to collect antlers from a group of individuals with knowledge of who owns each set, and examine how they vary within individual across years... But, if you had access to the antlers could a genetic identity test be a more reliable way to identify? –  GriffinEvo Apr 16 '13 at 6:37
1  
@violadaprile collecting dna as a form of ID can be done from living animals thus I am not advocating killing or physically interfering with a living animal. DNA can be collected from, amongst other things, faecal matter or antlers which have been shed. And I know the problem is how to recognize a deer, I raised the point of "do antlers vary between years within an individual?" and to answer this, and thus show whether deer can be identified reliably by antlers, would require comparing antlers produced within individuals. The choice of recognition method would depend on the observers needs. –  GriffinEvo Apr 17 '13 at 16:57

2 Answers 2

up vote 3 down vote accepted

There is a lot of variation in how and when deer shed their antlers. In most arctic and temperate-zone species, antler growth and shedding is annual, and is controlled by the length of daylight. In tropical species, antlers may be shed at any time of year, and in some species such as the sambar, antlers last several years. Some equatorial deer never shed their antlers.

The horns change every year and, especially, increase the number of branches (and consequently, change their shape). You can't recognize them by antlers, but by other features, such as color of the hair or the lineaments. Like us, animals have individual morphological differences that are recognizable and listable.

Biologists specializing in studies of particular animal species not only take photos, but also make drawings and write descriptions of behavior, to identify individuals within herds.

An optical examination, however, of the subject through drawings and photos (and if possible, direct observation), is more useful than a PC program. This involves identifying particular similarities and equalities that are not "identical". This is possible to do visually on a large (but limited) number of specimens. The human eye is the best computer.

share|improve this answer
2  
I wouldn't say that automated computer image analysis should not be used to study animals. Its hard to believe that at some point it will be more common than direct observation. –  shigeta Apr 17 '13 at 14:42
1  
it was just what i meant - you are biologists and microbiologists and informatics, i am only lover of nature and studied a lot about it. So we talk from 2 true different points of view –  violadaprile Apr 18 '13 at 0:54

The Open Source Computer Vision library OpenCV is pretty popular. I'm a Python guy, but it also has C, C++, and Java interfaces. The O'Reilly book Programming Computer Vision with Python was pretty good, and their C-oriented Learning OpenCV from 2008 is coming out with a new C++ edition in July, supposedly. There are also the OpenCV online docs, linking to the reference manual, user guide, tutorial, and Java API for each release.

share|improve this answer
    
MATLAB also has a computer vision toolbox –  WYSIWYG Apr 16 '13 at 3:27
    
Sure but... this implies that you can distinguish a deer by its photo, which is what we are trying to determine. –  nico Apr 17 '13 at 14:36
    
We have programs which use a tecnic knew as biometria. that presume an archive of thousands (millions?) of possible variants. I do not think that such an archive exists also for deer. –  violadaprile Apr 17 '13 at 14:51

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.