Group

Camera Traps / Feed

Looking for a place to discuss camera trap troubleshooting, compare models, collaborate with members working with other technologies like machine learning and bioacoustics, or share and exchange data from your camera trap research? Get involved in our Camera Traps group! All are welcome whether you are new to camera trapping, have expertise from the field to share, or are curious about how your skill sets can help those working with camera traps. 

event

Wildlife Crime Tech Challenge Accelerator Bootcamp

Sophie Maxwell
Earlier this month, the 16 prize winners of the Wildlife Crime Tech Challenge were called to Washington D.C. for an Accelerator Boot Camp. Sophie Maxwell, a member of the prize winning team from the Zoological Society...

0
See full post
discussion

Camera-trapping best-practices

Hi everyone,  As part of a WWF-UK conservation tech initiative, I'm putting together camera-trapping best-practice guidelines, aimed at field...

19 0

Hi Ollie

As a newbie to camera trap data, this makes interesting reading.

Just to check my understanding on terminology, in Figure 5, the reference to "Automatic filtering of blanks" - did that refer to eliminating false positives, so reducing the number of images that are not of interest ?

Andy

See full post
discussion

Discussion: Opinion of TEAM network and Wildlife Insights

So we have this camera trap dataset with all kinds of problems, and so we've been developing software that identifies the problems with the files prior to any kind of analysis...

5 0

Thank you. I had not seen TRAPPER.
I had seen Snoopy, CameraBase, the Sanderson & Harris executables, and TEAM. And none of them had seemed suitable for the work that I was doing.
I will investigate TRAPPER further.

I like TRAPPERs video capability.
I am yet to test it out though.
 

Hi Heidi,

You might also consider becoming involved in the eMammal project (emammal.si.edu) that the Smithsonian has recently developed. It has tools for both uploading and coding camera trap images, with the bonus that they are ultimately archived and made available by the Smithsonian for future researchers. My opinion is that moving toward open-access data will make all of our efforts more valuable into the future. I'm using eMammal, but not one of the developers or associated with the Smithsonian.

Cheers,

Robert

See full post
discussion

Resource: Wildlife Speed Cameras: Measuring animal travel speed and day range using camera traps

Travel speed and day range are metrics usually derived from telemetry or direct observations. A new (open access)...

2 0

Here's a set of tools that could be applicable to this idea

 https://github.com/pfr/VideoSpeedTracker

Hi Steph - just to follow up on your post: @MarcusRowcliffe , James Durrant and I have been working on a bit of software to implement the "computer vision" techniques that are mentioned in that paragraph. You can see a demonstration of it in action here. It requires camera-trappers to "calibrate" their camera traps during setup (or take-down), by taking pictures of a standard object (for example, we use a 1m pole held vertically) at different distances. The calibration takes ~10 mins per location. From this, you can reconstruct the paths that animals take infront of cameras, the total distance they travelled, and therefore their speed.

See full post
article

TEAM Network and Wildlife Insights

Eric Fegraus
Operating the largest tropical forest camera trap network globally, TEAM Network has accumulated over 2.6 million images. How can large datasets coupled with new techniques for data management and analysis provide...

0
See full post
discussion

[ARCHIVED EVENT]: Approaches to Analysing Camera Trap Data

On Friday 22nd April, the David Attenborough Building Stats Club will be meeting to discuss approaches to analysing camera trap data. This is a broad field, and camera traps...

2 0

Hi Steph,

Only just discovered this site, so I'm a bit late to the game.

I'd love to hear what your main take-aways were from this meeting!

Best,

Louise

Hi Louise, 

Welcome! Unfortunately, an uncomfortably busy calendar meant I ended up missing this gathering. However, I'm sure that @SteffenOppel @Tomswinfield  or @ali+johnston (I think you were all involved?) might be kind enough to jump in here and share some of their key take aways from this discussion?   

Steph 

 

 

See full post
discussion

Article: Google's cloud vision for automated identification of camera trap photos

I did a quick evaluation of whether Google Cloud Vision can be used off the shelf to automatically identify animals from camera trap photos. I wrote a blog post about it...

5 0

An update to the automated species identification debate: 

A paper has recently come out which used deep learning ("very deep convolutional networks") and managed 89% accuracy for the Snapshot Serengeti Zooniverse dataset, IF the image was first manually cropped around the animal. Seriously, who has time to do that? If the image remained uncropped they managed a woeful 35% accuracy. 

Perhaps we have a long wait ahead of us for this to become a practical reality?

Fair point, it isn't a peer reviewed article as yet. I had a poke around out of curiosity and wasn't able to track down a fully published paper yet (thought it's been a while since this preprint was released...). On reading, were there any issues that stood out to you that others should be aware of? And, as Ollie pointed out, the research wasn't getting a great response rate from uncropped images - I wonder if perhaps, given the time since the original publication, the results may have actually improved by now? It seems (from my interested but unqualified observer perspective) that the field is moving forward in leaps and bounds, such that a paper pre-published in march 2016 may very well be out of date by this stage...

See full post
discussion

Project Feedback Wanted: Building Low Cost Cameras

We have been facing some problems for Cameras either for monitoring nests or for quick deployment for medical or behavioral reasons. We need them to be low cost and easy to use...

2 0

thanks for this very interesting post! i was also trying to develop an inexpensive camera trap but with not good results. i think the use of a PIR sensor can give more battery life than motion detection via software. Can you give more details about the components you used please?

Thanks
Paolo

Hi Sorry been away, I'll list more about parts etc.. In the mean time the Pi Zero has just had an upgrade..

 

http://petapixel.com/2016/05/19/5-raspberry-pi-zero-now-camera-compatible/

See full post