Hi everyone,
We are about to start processing photos from a camera trapping study. There are many many thousands of photos and I cant help but think there must be some neat programs out there to speed up or automate the process. So far, 'Camera Base' (CamBase) appears to be the wildlife ecologists' tool of choice. Anyone out there have any other recommendations?
UPDATE: The identification of species will be done by eye (human). What I am looking for is suggestions of programs that would facilitate the process of going through many photos and that would generate a database based on our identification.
Thanks a lot!
Chris
So, to clarify: we will be doing the species identification by eye. However, I am looking for something that will aid in the handling of the photos and generate a database accordingly.
Will check these out though, thanks!
**EDIT: Sean, i just checked out the Serengeti project, VERY nice! Thanks for pointing that out.
Depending on how elusive or clear you species are in images, Image J can work really well to automate using landmark analysis
Hi Chris. I'm finishing up software called "Snoopy" for cataloging media (photos, audio, video and arbitrary data) for capture-recapture surveys. It is a research project to replace CameraBase, used by the researchers here at NSU. I will be making it available for Beta test early next semester. I've tested with 60,000 photos and it handles these volumes quite well. It's written in Java and will run on Windows, Mac OS and Unix. The database backend can be any ODBC compliant database. ve tested with 60,000 photos and it handles these volumes quite well. We are using MySQL as it is very fast and free. Here is a link to a Prezi presentation I did for the Oklahoma Academy of Science in November. http://prezi.com/xqoogmni0ymu/?utm_campaign=share&utm_medium=copy There's no narration for the online presentation but it will give you an idea of its capabilities. The software is free and will be available to anyone who wants to use it. Let me know if you would consider Beta testing. I'll be happy to answer any questions!
Hi Richard,
Terrific stuff, nice job with this and very nice Prezi!!
I will be in touch directly - I think we will definitely be interested in beta testing.
Chris
Great Chris. Feel free to contact me at your earliest convenience. I'm testing over the winter break as well, so Snoopy should be ready for prime time early in January.
Hi Chris,
CameraBase is definitely worth getting to grips with - it can be a lifesaver but there are a few glitches here and there!
Richard, "Snoopy" sounds and looks much more powerful / useful - I would be very keen to receive any updates too.
Cheers,
Louis
Hi Chris,
Just my thoughts. I have been busy with camera trapping for last 8 years and have found that Camera Base is still the best platform to storage camera trapping data. While it has its limitations (e.g. windows/access etc.) there is still in my opinion nothing which is better for our needs. Once the data is entered it sits in an Access database which you can access by R and allow for rapid data extraction.
Regarding your question on automated ID; this largely depends on the quality of the images. Unfortunately most digital camera traps have some motion blur, side angles shots which make automated ID difficult. That said, a colleague of mine have used Wild-ID (http://dartmouth.edu/faculty-directory/douglas-thomas-bolger; http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0059424) and he found that it correctly ID around 96% of leopard images. However, he used white flash camera traps which had very good quality images. So, it largely depends. For most infrared/black flash cameras my experience has been that manual ID is probably the best. I am looking forward to some alternatives, but all my data now sits in Camera Base and it will take some persuasion to move me to a new setup. Hope this helps.
Cheers.
Lourens
.
There is no any proper software which will identify individual marked animal. It is time consuming to identify each individuals by human eye. Once it is done, You will be very much confident while analyzing your capture re capture data in DENSITY or SPACECAP. On which species you are working?
There is some very good software for organizing camera trap photos available here: http://www.smallcats.org/CTA-executables.html
Hi Matthew,
Great, thanks for this. Looks like this is a good option for smallish numbers of photos. I wonder whether you find the need to do a lot of 'draggin 'n droppin' a little inconvenient? Looks like both CamBase and Snoopy provide a GUI that will do this stuff for you?
I am compiling a list of programs, and this will be on it for sure!
Thnaks
Chris
Snoopy has batch inport capability as well as batch capture and single capture selection. For example, you can import several thousand photos and either choose to capture all or capture specific photos (or other media). CameraBase has similar functions but is not as flexible as Snoopy's interface.
EMammal.org is a good option for species identification, the software streamlines the process of species id and data can be output in .csv format.
Arielle, thanks.
By the way, that link doesn't seem to go anywhere, but I found the site. Will figure out how it works and update.
Chris
It's still in development for outside use but that update should happen soon. Feel free to message me directly and I can get you connected with the team.
I can suggest this recently developed software, described in the following paper:
Krishnappa, Y.S. & W.C. Turner. 2014. Software for minimalistic data management in large camera trap studies. Ecological Informatics, 24:11-16.
I have two students who have started using it; how many photos are you actually sorting? I have a student with 130,000 pictures, so it's taking a while to even sort out grass from animals.
Hi Chris: As M. Butler wrote, a method pioneered by J. Sanderson has the following steps A) automatic renaming of images by date and time that the image was taken B) the user sorts camera trap images into a standardized file structure based on species names and # observed in the image and B) a set of programs analyzes this structure to make a text file describing species observed, location and time, plus a number of summary metrics. A number of Federal lands in the SW USA, and projects globally have been using this method successfully. It's been a huge time saver (especially for making occupancy matrices), and facilitates meta-analyses (by sharing and analyzing the text files, not the images themselves). We've tried to remove as much human input as possible (to reduce error), though the drag and drop does take time. To our knowledge, there are no automatic methods to identify species or # of them yet from such imagery. We've been handling millions of images this way. Literally - one site we analyzed had > 2 million images.
References include: Publication - http://www.esajournals.org/doi/full/10.1890/0012-9623-91.3.352
Publication- http://www.smallcats.org/files/Sanderson_Harris2013.pdf
New book - http://www.publish.csiro.au/nid/18/pid/7150.htm
Programs and further material: http://www.smallcats.org/CTA-executables.html
Hope you find this useful.
Grant
Hi Chris, just in case, a simple way of extracting exif information and putting it into R is to install exiftools and call it form R with *sytem* comand
Thanks, dear colleagues. Most possibly we will need mentioned software. So far working by hand, but some cameras produce a lot of nonsence (cars, for example). I think, that over 10 000 pictures per month definitely should be at least sorted by a program.
Interesting Timothy. I wrote a similar program ("YakTalk") over the summer that automatically extracts audio for analysis. I had 6 TB of starling chatter I needed to extract for analysis (syntactic/grammatical). However, I was planning to include a module for video extraction. I will definitely be looking at motionmeerkat. Thanks for sharing.
Hi! There is going on "Wildlife Photo-ID Network" workshop and you find the keynote speaks here: https://www.youtube.com/playlist?list=PLaNfvlZd-a3UarXVi4RqDxNIbo71Kxl7G. Talk one introduces Discovery software (http://www.biosch.hku.hk/ecology/staffhp/lk/Discovery/) and talk four IBEIS (http://compbio.cs.uic.edu/IBEIS/). Here is also a link to the first newsletter of the workshop where some photo-ID related softwares are shortly indroduced (http://www.uef.fi/documents/2317961/2334570/NEWS_letter_1.pdf/51f7733a-14bf-4c82-8122-4a033142d608).
Thanks very much Miina,
Great resource - just having a look to see whether this is as good when you don't want to ID individuals as it is for when you do.
Thanks...
Take a look at the TEAM network software for this: "deskteam". It is free and provides a number of useful functions for annotating and batch processing. It has been used at a number of sites (producing a sum total of nearly three million images so far) for the last five years so should be pretty bug free.
See http://www.teamnetwork.org/help-deskteam
Also some background information http://www.teamnetwork.org/help-deskteam
We've used the free programs described here: http://www.smallcats.org/files/Sanderson_Harris2013.pdf
If you are sorting by eye, this provides a way to analyze the findings thoroughly once you get the hang of it.
Thanks R. Smedley for your answer and update about the software. I am also having a stock of more than 15 - 20 K pics every year from camera trapping exercise. He be great if you can consider for beta testing.
Hi folks,
The Beta version of Snoopy will be available here soon. I've created a blog and message board on SourceForge.net. Currently, I'm working diligently on the capture and offset portion of the code, and it's looking good so far. Here's the link where the beta version will be available. https://sourceforge.net/projects/snoopy-capture/
This might be of help:
Camera trap data processing manual, by Jim Sanderson
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CCgQFjABahUKEwjHz5D_p5fHAhUBdj4KHU9fDBs&url=http%3A%2F%2Fwww.smallcats.org%2Ffiles%2FCamera_trap_data_processing_manual.doc&ei=KtLEVceSK4Hs-QHPvrHYAQ&usg=AFQjCNEj_Lfc9Nz5IuC6oreR5pW4JjjuxQ&bvm=bv.99804247,d.cWw
Hi R. Smedley,
A pitty that the link doesn't work (http://sourceforge.net/projects/snoopy-capture/). Can you check the link?
Thanks for that! Otherwise I need to try CamBase.
René
Hi Rene,
I moved Snoopy to its own web site. Go to http://www.tulsasoft.com and click on the menu selection "Snoopy".
Regards,
Richard
Hi Richard!
This looks REAL helpfull for my study on wild cats in the Netherlands! I download is, but... How do I become on a UserID and Password?
Thanks for making and sharing this (looks like) helpfull program!
Rene,
Watch the tutorials on installation. You have to install database software first (SQL Server for Windows or MySQL for the Macintosh). You will set you user id and password in the database software. SQL Server installation is fairly involved, but I go through it step-by-step in the tutorial. I will eventually create an automated install program, but that will be in a later Snoopy version. Feel free to email me directly at [email protected] or Skype me at smedley30. Be sure and let me know who you are if you Skype and request contact.
Regards,
Richard
Hi Chris. I'm wondering if you might provide an update for this post - which software did you end up using and how is it working for you? I am about to begin processing photos from a camera trap study and have questions that are similar to your original post.
Thanks very much,
Katherine
We went with camera base in the end. http://www.atrium-biodiversity.org/tools/camerabase/
We may have gone with Snoopy becasue it sounds great, but it was in beta stage when we needed it so we didn't go down that route.
I am toying with the idea of setting up a google site/listserve as a central place for these types of discussions.
Good luck!
Chris
There is a nice article in Wildlife Professional (attached) that mentioned a couple of different software options, but the lack of good full fledged options is a problem across the profession. In addition to our own in-house data management, we are looking for a online platform for a citizen science program where the user uploads, IDs and submits their cameras trap photos to us. A basic mapping function would be a bonus. Just got a quote for $25,000 to develop which is outside of our budget! Nothing seems to exist along these lines...any suggestions are welcome!
I released Snoopy Version 1.0 last week -- http://www.tulsasoft.com. Got some heavy beta testing from some folks studying Jaguars on the Amazon and was very helpful ironing out the bugs.
Hi Christy,
This is a really cool site for doing some of the things you are looking for (I dont know how much it costs, but I like it nonetheless):
http://www.snapshotserengeti.org/
Hey Chris, thanks for the suggestion. The Zooniverse program is really powerful but it is a crowd sourcing, data generating platform that costs a lot (quote of $200,000 for a custom project). We are looking for something that our camera users can use to annotate and share their own photo data with us and we need privacy settings (it's a community species richness study and the cameras are in their back yards).
Loving all the feedback and ideas on this question!
Hi Richard,
does it already works on a online database? To make an own database, it doesn't work easy for me and not so easy to work with different people on it.
Christy: An other option is to join the Cameratrap database of WUR: http://cameratrapping.net/. They ask a bit money for datatransfer and space...
René
By the way, in addition to the Wildlife Photo-ID Network and other sources mentioned previously there is a Yahoo discussion group called 'Camera Trapping Information Exchange': https://uk.groups.yahoo.com/neo/groups/cameratraps/info
Russ,
Not sure why I hadn't come across this earlier! Thanks for sharing.
Chris
Hi Rene,
You need to install either MySQL (Mac/Unix) or SQL Server (Windows) for Snoopy. I'm working on H2 (another, self-installing ODBC database) for future versions.
Richard
Kamiel Spoelstra added his self written Software SpeedyMouse to ResearchGate: https://www.researchgate.net/publication/289202434_SpeedyMouse_22_for_the_analysis_of_camera_trap_images
It works easy!
Research SpeedyMouse 2.2 for the analysis of camera trap images
Hi Richard, I'm going to try your Snoopy software with our camera trap data. Do you have a mailing list for new updates or something?
Thanks
Hi all,
If somebody looks for a web based and open source solution here is our proposition.
Best,
Kuba
https://demo.trapper-project.org/
https://demo.trapper-project.org/forum/
https://bitbucket.org/trapper-project/
Article TRAPPER: an open source web-based application to manage came...
Jakub,
I just finished looking at the paper - nice work, and thanks for sharing.
Chris
So TRAPPER and camtrapR are now both published in Methods in Ecology and Evolution, great stuff, thanks to the authors for all the great work.
I have a couple of observations (questions perhaps):
TRAPPER:
http://onlinelibrary.wiley.com/doi/10.1111/2041-210X.12571/full
is it the case that I can only use TRAPPER using Ubuntu as the operating system. The documentation alludes to there being the option of getting up and running on other platforms, but no instruction is given. I am thinking about this from a usability perspective and teaching undergraduates about camera trap data management in particular. for example, i sent some student that were very familiar with camera base to 'start a small trapper database' using a subset of the data and they struggled to get going.
camtrapR:
http://onlinelibrary.wiley.com/doi/10.1111/2041-210X.12600/epdf
I have 100's of thousands of photos. The species ID process is indeed laborious as the authors say, and has to be done externally to camtrapR. The neat thing about Camera Base and TRAPPER is that you can do the species ID, or 'tagging', **and the relational database get generated simultaneously**, from which, using queries, any type of data format can be generated. Then you need to read into R, but this is easily automated, especially using R-to- functions (I think they exist). So, which two-step approach is better? I don't know?
(***in very early stages of tinkering with all of these so apologies for any inaccuracies - and VERY happy to be pointed in the right direction***)
I will continue to sporadically provide updates, and would be keen to hear about others' experiences.
Chris
Hi Chris et al.,
camtrapR is indeed a very welcome contribution, great stuff, thanks the authors for that! As Chris has suggested, both applications seem to be compatible and when used in a two-step approach can significantly improve the workflow in cam-trap projects. I think there is a serious potential in enhancing this two-step approach/path by creating a link (e.g exchange of data through the API in TRAPPER) between these two open source solutions.
Marcin
One of the main ideas behind TRAPPER is that it is a web application i.e. it has to be installed at the server of your institution, department etc. I understand that this step might require IT expert knowledge, but once TRAPPER instance is running, then your students and everyone else can directly use it. If this first step is a problem for you we can help you. To simplify TRAPPER setup we are working on a virtual machine which will contain already pre-configured web server and all the other stuff needed to run your own TRAPPER instance.
I fully agree with Marcin that this is a great idea to think more about how both solutions (camtrapR and TRAPPER) could be combined together!
Best,
Kuba
Marcin,
Yes! Integrating (1) the datailed database creation with (2) general and flexible R-based data processing and analysis, is indeed the golden ticket.
Jakub,
Thanks for the offer. I will investigate the availability of server space and be in touch. I do think of smaller institutions and organisations, and those conducting studies in less developed countries or on less devloped projects and think this initial step may be limiting, so the idea of a virtual machine is a great one, and one I hope works out.
(By the way, i think these developments are terrific and a very welcome contribution to the camera trapping community, and offer only [personal] opinions on my experiences as hopefully constructive feedback)
Chris
Hi Chris,
No it is a software, but an australian government funded use a particular method for processing thousands of photographs of trap cameras . To address it , they have designed a very curious collaborative website where citizens are asked . His website is:
https://www.zooniverse.org/projects/birgus2/western-shield-camera-watch/classify
Perhaps in society united states also have acceptance of this methodology.
Regards,
Jesús A. Cuevas
https://www.zooniverse.org/projects/birgus2/western-shield-camera-watch/classify
Just thinking outside the box. Have you given thought to using something like Zooniverse (https://www.zooniverse.org) to get some citizenscience help in IDs of your photos. Here is an example on which I have helped (the contributors seem to range from amateur naturalists to crusty ol'zoologists like myself) https://www.zooniverse.org/projects/birgus2/western-shield-camera-watch/.
Hi all,
Just a small update:
The pre-configured TRAPPER Virtual Machine (Ubuntu Server 16.04 LTS 64bit) is available for download at:
http://dx.doi.org/10.5281/zenodo.57807
More info at:
https://demo.trapper-project.org/forum/topic/5/
Best,
Kuba
Hi all,
Im using this set of programs to analyze database of camera traps. This programs run in DOS and generate results in file TXT. This TXT file is easily attached by email.
This program was made by Jim Sanderon, you can visit this http://www.smallcats.org/CTA-executables.html also is recommendable read this paper
http://www.smallcats.org/files/Sanderson_Harris2013.pdf.
Previusly i use CamBase, now im try to learn use camtrapR package, but i dont have enough experience in R. I think that camtrapR is very powerfull tools to analize cam trap data base and share several similarities whit sanderson program.
Regards.
Hi everyone,
I wanted to let you all know about yet another option for camera trap photo management called Camelot:
https://gitlab.com/camelot-project/camelot
In the coming months I will also share info about 'vixen', a python based tool that appears to manage video seamlessly as well as stills.
Enjoy!
Cheers Chris. We are still waiting on decent software for video management - Vixen sounds promising!
So there is a camera trap discussion group here: https://www.wildlabs.net/community/group/camera-traps
I worked with Chris Mann, the developer of Camelot:
https://gitlab.com/camelot-project/camelot
And we have a Google Group here:
https://groups.google.com/forum/#!forum/camelot-project
TRAPPER is another tool I'm really looking forward to try as soon as I get back to the city and have a proper connection!
Article TRAPPER: an open source web-based application to manage came...
We are looking for people interested in reviewing our paper for Camelot. We are submitting it to Methods in Ecology & Evolution. Any offers?
It's a 3000 word application paper.
Hello everyone,
Another really flexible software/tool kit that should of interest to followers of this thread: ViXeN: An Open-source Package for Managing Multimedia Data
Chris
http://onlinelibrary.wiley.com/doi/10.1111/2041-210X.12892/full
Hello all,
A new software for camera trap analysis
https://www.zsl.org/zsl-camera-trap-data-management-and-analysis-package
Rosa
Hi All,
You might find some useful guidance on this and other camera-trapping questions here: DOI: 10.13140/RG.2.2.23409.17767
There are some webpages and other guidelines available here if interested: https://www.wwf.org.uk/conservationtechnology/
Best wishes, Paul
We now have a paper on Camelot available:
Our paper: https://www.biorxiv.org/content/early/2017/10/18/203216.article-info
WWF mentions:
https://www.wwf.org.uk/conservationtechnology/camera-trap.html which takes yo to the full guidelines:
https://www.wwf.org.uk/conservationtechnology/documents/CameraTraps-WWF-guidelines.pdf
We had to review tens of thousands of camera trap images and we have been using the free Timelapse2 Image Analyser (http://saul.cpsc.ucalgary.ca/timelapse/ ). Besides being simple to install and use, the advantage that we found in this program over a few alternatives that we tested was the magnifying glass that it provides and you can point to the parts in your image where you need to enlarge and see small details. Not every one needs this tool, but in our images the animals that triggered the camera were often not very close and could be quite camouflaged (Uromastyx lizards on a diverse stony background) and the magnifying glass was extremely useful.
Hi Chris you will find useful information in following review: https://www.researchgate.net/publication/327503296_Software_to_facilitate_and_streamline_camera_trap_data_management_A_review
Hi people... I am so glad that I followed this question on 'Software for processing and collating camera trapping photos'. After read some answers I started trying few softwares. I work with camera traps for over 10 years and I was struggling to process and edit all information from videos and photos. I am already using Timelapse2 (http://saul.cpsc.ucalgary.ca/timelapse/pmwiki.php?n=Main.HomePage) for few days and I am positively surprised with the program. It helped me to speed up the job and is really good at reading date/time as well as other really cool things. It seems to be all I needed for my camera trapping data.
maybe it's useful
http://www.atrium-biodiversity.org/tools/camerabase/
Cheers