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. 

discussion

Data Needed: Camera trap images for machine vision training

Hi everyone, I'm working on an object detection system for detecting elephants! It's detailed here https://www.wildlabs.net/community/thread/284 and at my project...

8 0

Hey, this is really interesting @theraido, thanks for sharing. Kirk has a nice writing style that's really easy to understand.. to the point where I feel lulled into confidence that I'd be able to replicate the project without too much fuss (I suspect this might be rather a strong case of over confidence). I was also amused that waiting for the birds to appear at the feeder seemed to be the slowest part of the project. 

It looks like it's going to be a series of tutorials, with the goal being to build an open platform for exploring crow intelligence. Eventually, he says like to identify individual crows in the wild, and give them the opportunity to pass a series of tests. So it'll be interesting to follow along and see whether he reaches this goal. 

In case anyone is interested, the next tutorial posted so far is 'building the rich man's deep learning camera', where he shares how to build a self-contained deep learning camera to detect birds in the wild.

 

Sounds great!

I am doing the almost same thing, since here we have a problem between elephant and human living in the same area and conflict a lot. We would like to warn people when elephants are approching by camera trap's detection of elephants.

How is your work now? Maybe we could share some info and put it forward together.

Anton

See full post
discussion

Recommendations Needed: Software for sampling photos from video

Hello Camera trappers :) I am using video in our cameratraps sampling but need to extract picture from them to identify jaguar individuals. Any suggestion of good software to...

6 0

My preference is to use Adobe Lightroom, allowing me to move forwards or backwards one frame at a time. I can also adjust the exposure. This is an example from one of my videos. 

I hope this is helpful. 

 

See full post
discussion

Camera Trapper from Colorado

I'm honored to be part of this WILDLABS group and specifically this Camera Trap forum. I hope that what I've learned weekly over the past 14 years will help others. Most...

0
See full post
event

Camera trapping workshop, London

Amy Fitzmaurice
This one day workshop will cover key aspects of planning, executing and analysing camera trap surveys in the field. It will also examine the various survey techniques for which remote cameras provide a powerful tool and...

0
See full post
funding

The Plant-Powered Camera Trap Challenge

Alasdair Davies
Are you an architect, engineer, designer or a scientist? Can you design and manufacture a prototype open source plant-BES (bio electrochemical system) to power a camera trap and environmental sensors in tropical forests...

0
See full post
article

#Tech4Wildlife Photo Challenge 2018: Our Top 10

WILDLABS Team
Hundreds of people joined our #Tech4Wildlife photo challenge this year, showcasing all the incredible ways tech is being used to support wildlife conservation. We've seen proximity loggers on Tasmanian Devils in...

0
See full post
discussion

Research: A rigorous, realistic and reproducible test of camera trap perfomance

I am working towards having a test of camera trap performance where the targets are real, moving animals instead of people, the movement of the targets is controlled so that tests...

5 0

Also, just a general comment: some less epxensive cameras peform very well but may be more prone to 'glitches' over the duration of an extended study. I think the expectation of long term reliability is part of the reason some people choose the expenisve brand. Systematic tests of long duration reliability in field conditions would be really interesting, albeit probably too difficult/expensive to achieve.

Thank you Julia

Camera testing is certainly not Toffee's favorite activity - he would much rather be sniffing for scent marks !

The adverse effects of high temperatures on PIR are well established and they are a major problem anywhere that air temperatures get above about 30C. There is also a problem with cameras staying hotter than their surroundings for a few hours after sunset. I have also noticed that the infrared illuminator on the Reconyx actually heats up the camera.

Birds might be trickier to train than dogs, but you only need a reliable way to lure them to particular points within the field of view.

Certainly there are more factors to consider than only detection capability (though arguably that is the most important - better a fuzzy picture than none at all probably) and reliability is one of those. Bushnell Trophycams are notorious for losing their date settings (and this morning the one I am testing had done just that) for example. All sorts of equipment gets put through accelerated durability tests, and there is no reason why camera traps should not be similarly tested. 

Given the huge projects that are built around camera trapping, and the scale of the conservation management decisions that are based on camera trap data it is a real problem that their performance is not tested and validated as fit for purpose.

Peter

The sequence of images in the attached brief report shows why camera traps must be tested with real animal targets, and not with humans. The camera easily detects a human, but misses multiple images of the target dog.

See full post
event

Technology Empowered Conservation Lecture Series

Paul Jepson
New technological forces look set to transform biodiversity science. This series will showcase and discuss cutting-edge applications happening in Oxford and beyond. It is guaranteed to inspire and challenge. 

0
See full post
article

Instant Detect 2.0: A Connected Future for Conservation

Sam Seccombe
‘The Field’… Say the words ‘The Field’ to a group of conservationists and it will immediately conjure up vivid memories of everything from sticky wet rainforests to burning dusty deserts. What’s more, it’s almost...

0
See full post
discussion

Resource: Camelot - new camera trap software

Camelot is an open-source, web-based tool to help wildlife conservationists with camera trapping. Camera Trapping software for Species Identification & Reports. By...

14 0

Hi Egil, I think Camelot has a reasonable story around most of these things. Here's how I see it:

I set up a survey and I add the information about all the camera trapping stations (including camera IDs).

I import the photos. The database reads the data, time and camera ID from the photos so photos are linked to a camera trap location.

Camelot has two modes for importing data: a bulk import, and a per-session import.  It sounds like you have many images up-front, and so bulk import may be the way to go.  In this case Camelot can create the camera trap stations based on the location of the images, and their metadata.

http://camelot-project.readthedocs.io/en/latest/bulkimport.html

Then I do a quick first pass where I assign a species (tiger, leopard, etc) to each photo.

Yes, I expect you'll find the library UI suitable for this.

In a second round I want to identify the tigers. So I get all photos labelled with 'tiger'. Obviously the first one is a new animal thus I want a 'button' which allows me to add a new animal. Then in following pictures there is a drop-down menu with identified tigers. The order of the pictures presented is by camera trap location, data, time and then the next nearest camera trap location.

The sighting added in the first pass can be edited to add-in the individual, and new individuals can be added straight from the dropdown menu.  Camelot does not have the ability to define a custom ordering of images: the ordering is always by camera trap station, by capture time (broadly; it's a bit more nuanced than this in reality). However does have the notion of "reference images", where images flagged in this way can be displayed, in another window (which can handy where multiple monitors are available), based on the currently selected sighting field data.

If things go as planned I'll likely be running into 100,000's of pictures, so I need to be able to do the first round pretty quickly. Then there will be 4-5 species I need to identify at the level of the individual. There will probably up to about 200 individuals of a species, but never more than about 20 at each camera trapping station, and the vast majority of individuals won't show up at more than 4-5 camera trapping stations either.

Camelot does support this sort of raw scale (x00,000 images) and has a reasonably efficient UI for identification.  The information here may be relevant, depending on how many multiples of 100,000 it turns out to be.

However Camelot does not currently offer the ability to limit individuals in a field depending on the selected species (it is always the same individuals available in a dropdown regardless of the selected species).  Potentially this limation could be worked around by having a dropdown available for each species, which will at least highlight in an export where a data-entry error may have arisen.  (e.g., individual chosed in the "individuals" for species X field, but the selected species is Y.)

At 200 individuals the workflow for reference images could start to break down too, depending on the level of familiarity with the individuals. (i.e., repeatedly searching through all reference-quality images of a species, or images for a couple dozen individuals to make an identification could be onerous.)

I've looked around and I think Camelot is closest to what I want, but I wonder if I could get it really close to what I want. I have quite a bit of experience with SQL and building access databases, but I thought, with the prevalence of camera traps these days, that there would be more packages out there.

I agree Camelot really only goes part-way to meeting the requirements.  It should be workable to use Camelot for the purpose, though support for individuals is relatively new and really hasn't been optimised for this sort of scale. It seems like the two places where Camelot is most lacking for this workflow are (and correct me if I'm wrong):

  1. lack of some more flexible ordering system for reference images against the selected image(s)
  2. lack of ability to filter options for a single "individuals" field based on the selected species

It'll likely be some time before these features are available in Camelot, but I'll add them to the development backlog.

-Chris

Thanks for the reply Chris!

When you state:

"Camelot has two modes for importing data: a bulk import, and a per-session import.  It sounds like you have many images up-front, and so bulk import may be the way to go.  In this case Camelot can create the camera trap stations based on the location of the images, and their metadata."

When you mention 'location' you refer to the location of the images from where they are imported from? And not the gps location from the EXIF data if it's available, right? In my case no gps data would be available. But if I've downloaded the images to a hard drive in the field, and then back in the office import them into Camelot I can tell Camelot that the pictures from folder X belong to camera station 1?

You're right about the two points for optimizing. If photos can be ordered by station, date, time, and reference images ordered by 'distance to station', it greatly reduces the number of reference images to look through as the vast majority of individuals will be recorded at only a few stations, effectively trimming down the number of individuals to look through from 100's to a few dozen.

Cascading the drop down lists Species > individuals (or even area > species > individuals, or species > area > individuals, or in some cases species > group > individuals) would be beneficial too in the process of identifying individuals. In a SQL database this isn't hard to code into the fields on a form, but might not be that easy in Camelot as you would have to chose if the field you add depends on another field on the form.

Hi Egil,

When you mention 'location' you refer to the location of the images from where they are imported from? And not the gps location from the EXIF data if it's available, right?

Yes, that's right -- I should have said the 'directory of those images'.

But if I've downloaded the images to a hard drive in the field, and then back in the office import them into Camelot I can tell Camelot that the pictures from folder X belong to camera station 1?

Yes, that's right.  Images can be dragged & dropped into existing camera trap stations, which have GPS coordinates associated.  For the bulk import case, the data scanned from the image would need to be joined with GPS data to produce the final CSV for upload.

And thanks for the confirmation on how you'd expect that functionality should behave.

-Chris

See full post
article

FIT Cheetahs

Larissa Slaney
A new research project is looking to investigate whether technology combined with the ancient skills and knowledge of Namibian trackers can help save cheetahs from extinction. Called FIT Cheetahs, the research project...

0
See full post
article

HWC Tech Challenge Update: Meet the Judges

WILDLABS Team
Our panel of international experts has been hard at work reviewing the 47 proposals we recieved for innovative technological tools to address human wildlife conflict. The panelists have systematically been assessing the...

0
See full post
discussion

Recommendations Needed: Best camera traps for Central African rainforests?

Hi all,   I have a question, I know someone who is looking at setting up camera traps in some somewhat remote parts of Central Africa. This would be primarily done...

1 0

Hi John, 

They will want a camera with good tolerance to humidity + precipitation (so a camera with a proper O-ring seal, and pack it with regularly-dried silica gel). Elephants will likely have a go at the cameras - so they will want to think about protecting their cameras with heavy-duty steel security boxes (perhaps with welded on spikes, which has worked in Thailand) and minimising smell left on and around their cameras during setup (e.g. use gloves, don't smoke or leave food).

Apart from that, the usual considerations apply: good detection circuitry (less of a problem if targeting elephants though, as they are massive) and battery life always helps. 

Laila Bahaa-el-din et al. used Panthera and Scoutguard in Gabon; the Goualougo Triangle Ape Project in Rep Congo uses Reconyx I think; Julia Gessner et al. used Reconyx in Rep. Congo and Cameroon. The latter reported that (of 47 cameras), 4 were taken out by elephants and one by a leopard!

Ollie  

See full post
article

How to lose a BRUV in 10 days

Ivy Baremore
As pressure on marine resources increases, fishers have to explore deeper and deeper waters to make a living. What does this mean for Belize’s deep-sea sharks? In an effort to understand the threats to these animals,...

0
See full post
discussion

Recommendations Needed: GSM Camera Traps

Hi everyone, I was wondering if people had any decent GSM camera trap recommendations We've been using Scoutguard and Spartan camera traps and have had success with...

5 0

Hi Chloe,

I've had promising results with the Scoutguard MG983G-30M, and I believe a 4G version has just been released. The 3G version has some useful features, like two-way communication to change settings/get images and a audio call feature (which I haven't used). I'd be very pleased to hear how your tests go.

Cheers,

Rob

 

Thanks everyone,

@TopBloke @Kai - I've not come across the Ltl Acorn ones - will take a look. So does that mean if it was purchased in the UK it would be locked to a UK SIM then? That might be problemativ as we tend to purchase here to test and set up before  sending out when staff are heading back to the country of deployment.

@Rob+Appleby - That's good to know about the Scoutguard MG983G-30M too, thanks. 

A supplier is going to send me a new model by the same people who make Scoutguard to test but he hasn't said what it is. Maybe it's the 4G one. Whatever it is if it is any good I will feed back on here

Chloe

@@Chloe+Aust -  pls google “ltl acorn uk”, you will get a lot info.  No, no limited, just the company's sale rules. You could use it anywhere, but for example, if you wants to use it in kenya, you need a local sim card, such as safaricom... when you check photos from china, you will find most of them photoed by are ITI acorn. Panda,tiger, snow leopord, cloud leopard, leopard, golden cat.... 

Scoutguard is also a good china camrea traps. oldest company. only company have espano menu. but they are focus on US hunting market. not special for wildlife researching use.

The LOREDA is also good, the only brand focus on wildlife-bio researching camra traps in china. Might be do not have dealers/stock in uk.

 

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 here....

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

Resources: Panthera Camera Trap information

Greetings Camera Trap community! I'm sure a number of you have heard about Panthera's camera traps. I thought to share two resources with you, one on Panthera's, I...

2 0

My understanding is that the Panthera cameras are only available to people or groups in Panthera's network. I tried to get some a few years back and had no luck, even for a project that had received Pathera funding in the past.

Regarding the poacher cam, I know @ColbyLoucks at WWF has developed a system using infrared thermal cameras that worked in the field trials.

Putting on my black thinking hat, the design as shown of the poacher cam will not be effective long-term once poachers know to look for it. Eric Dinerstein had worked on a project where such cameras would be concealed in a vine or some other organic-looking encasement. 

See full post
article

Machine learning, meet the ocean

Kate Wing
There is a revolution coming in conservation. Advances in conservation technology are generating more data than ever before on what lives where, who eats who, and what’s disappearing and how fast, but it still requires...

0
See full post
discussion

Camera Trap Pictures Wanted

Hi All, Does anyone have some camera trapping imagery they would be willing to share? Imagery will be used for best practice guidelines and supporting website, and will be...

7 0

Steph,

Not quite. I'm coordinating a fauna monitoring project for a Landcare network and, of course, we hope to get phascogales. I've just retrieved a batch of cameras from nest box monitoring duty for the phascogale project and their next role will be to investigate a possible sighting of a Squirrel Glider in the area.

Colin

See full post
discussion

Recommendations Needed: Real-time enabled camera traps

Hello all,   I have become interested in camera traps and the possibility of using them in remote areas. I was wondering what are the best reccommendations for camera...

9 0

Hi Jankees

Thanks for your reply,

If there have existing network like RoyalKPNN.V. provide mobile phone service,  there will be no problem, 3G/4G camera traps will works well. each cam 250 USD at the moment. 

Now days hundreds of chinese tech team are working on LoRa and NB-IoT solutions as well. All of them want to be successfull like HUAWEI,ZTE and DJI. Nice people, good team, team work ,working 12 hours each day like machine, like arms race.

I am sure they are willing to support anti-poaching. I am not techman,but we are chinese, we have the duty to solve problems, be responsible for it. If you come to china oneday, let me know.

Thanks, and your sensingclues is great. 

Regards

Kai

A number of good points have been made. In terms of remote-enable camera traps, you will mainly find one that use cellular data signals to transmit images. Typically these are thumbnails rather full resolution, so you will likely still need to retrieve teh emory cards for analysis. Also, traps that transmit images tend to have a lionger time lag between shots, which can be a problem.  Hunting web sites tend to have the most complete reviews and discussions of the various models.

If you want to use citizen science in the data analysis process, I would suggest looking at Zooniverse (www.zooniverse.org). It has a pretty well-thought-out platform. 

See full post
discussion

Survey: Camera trap effects on people

A team of Cambridge University researchers (Rogelio Luque-Lora, Bill Adams and Chris Sandbrook) is looking into the implications for people of the use of camera traps for...

1 0

Thanks, i send the it  to my chinese friends, i am sure some of them finished the survey.

cheers

Kai

See full post