Image recognition of deer in fields

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Image recognition of deer in fields

Nicolas Houde
Hi I am working in animal ecology,

I would like to do some image recognition of deer from pictures taken in fields.

I have not found the way to do it yet.

The difficulty is that the deer are pretty well concealed in the picture  as you can  see on the picture attached to this mail.

 

The light conditions are also pretty variable from one picture to the other.

If someone has a idea on how to resolve my problem, please let me know.

Thank you very much.


Reply | Threaded
Open this post in threaded view
|

Re: Image recognition of deer in fields

Rocco D'Antuono
This post was updated on .
Hi Nicolas, I made an attempt with Advanced Weka Segmentation plugin:
-find attached the segmentation after manual classification of the animal "Probability_maps-1";
-the second picture is an attempt of simulating different animals "PICT0746-mod";
-the third picture shows the application of the same classifier to the modified picture, as you can see, the identical copies of animal are found, but if brightness changes (like I guess for other photos under different illumination conditions), things become difficult...
You should try to segment several pictures, let me know if it works.
Best regards,
Rocco

Senior Microscopist
Crick Advanced Light Microscopy facility (CALM)
The Francis Crick Institute
1 Midland Road, NW1 1AT, London (UK)
https://roccodant.github.io/
Reply | Threaded
Open this post in threaded view
|

RE: Image recognition of deer in fields

Nicolas Houde
Hi Rocco,
Thank you for helping me with my problem,
I will try this plugin and I will let you know how it works
Were there any documents that should be attached to your last mail?
Have a nice say
Nicolas


Date: Mon, 10 Mar 2014 09:04:50 -0700
From: [hidden email]
To: [hidden email]
Subject: Re: Image recognition of deer in fields

Hi Nicolas, I made an attempt with Advanced Weka Segmentation plugin:
-find attached the segmentation after manual classification of the animal "Probability_maps-1";
-the second picture is an attempt of simulating different animals "PICT0746-mod";
-the third picture show the application of the same classifier to the modified picture, as you can see, the identical copies of animal are found, but if brightness changes (like I guess for other photos under different illumination conditions), things become difficult...
You should try to segment several pictures, let me know if it works.
Best regards,
Rocco




If you reply to this email, your message will be added to the discussion below:
http://imagej.1557.x6.nabble.com/Image-recognition-of-deer-in-fields-tp5006703p5006854.html
To unsubscribe from Image recognition of deer in fields, click here.
NAML
Reply | Threaded
Open this post in threaded view
|

RE: Image recognition of deer in fields

Rocco D'Antuono
Hi Nicolas, I posted my reply on the forum: there I could attach images, have a look at the page

http://imagej.1557.x6.nabble.com/Image-recognition-of-deer-in-fields-td5006703.html

Best regards,
Rocco
Senior Microscopist
Crick Advanced Light Microscopy facility (CALM)
The Francis Crick Institute
1 Midland Road, NW1 1AT, London (UK)
https://roccodant.github.io/
Reply | Threaded
Open this post in threaded view
|

RE: Image recognition of deer in fields

Nicolas Houde
All right!
Thank you very much
I will try this
Nicolas


Date: Tue, 11 Mar 2014 03:31:12 -0700
From: [hidden email]
To: [hidden email]
Subject: RE: Image recognition of deer in fields

Hi Nicolas, I posted my reply on the forum: there I could attach images, have a look at the page

http://imagej.1557.x6.nabble.com/Image-recognition-of-deer-in-fields-td5006703.html

Best regards,
Rocco



If you reply to this email, your message will be added to the discussion below:
http://imagej.1557.x6.nabble.com/Image-recognition-of-deer-in-fields-tp5006703p5006864.html
To unsubscribe from Image recognition of deer in fields, click here.
NAML