Video editing
Images are sometimes not shown due to bandwidth/network limitations. Refreshing the page usually helps.
You are currently reading a thread in /g/ - Technology
You are currently reading a thread in /g/ - Technology
![Video 108.mp4_snapshot_08.15.44_[2015.12.22_15.27.08].jpg](https://i.imgur.com/PtZtpf8.jpg "Video 108.mp4_snapshot_08.15.44_[2015.12.22_15.27.08].jpg")
Hi /g/
I'm working on a graduation project with weaned piglets. For this shit ass project we recorded 72 hours of video for 4 different pens. We marked all the pigs so we could differentiate them on camera. Now I have to wonderful task of observing the video's and scoring the pigs behaviour.
There's on problem tho. Halfway through the markings on the pigs are getting harder and harder to see. It doesn't help that the webcams recorded in absolutely shit tier quality.
What's the easiest way to edit the video's to maybe increase the contrast so the colors pop out more, and are better visible?
OpenCV, track pigs from the start automagically.
>>51997667
I have to use a program provided by my uni to analyze the videos. (Noldus Observer). So I can't use another program to analyze the videos.
>>51997704
Then you're not allowed to edit them either, my suggestion was to track and label them automatically for you to see easier.
But go ahead and upload a screenshot of the video to some online editing site and fiddle with contrast and colors to see if it helps then get video editing tool of your choice and apply the same to the whole video.
>>51997704
Then you're thoroughly fucked.
Do the recording again, but this time mark them by having them wear something. Like a hat or a robe.
It's just 3 more days anyways. And you'll get it just in time for Christmas.
>>51997759
Editing wouldn't be a problem long as I use the given program to analyze the pigs. I just need the markings to be more apparent so it's easier for me to follow the pig and it's behaviour.
And I'll go ahead and try that. Thanks.
It would be good if the program we have to use automatically tracks the pigs. Would save a lot of hassle.
>>51997776
Can't the pigs we observed have been moved to Germany. The stall we observed in was just to raise the piglets till they're about 30 KG's in weight. After that they get moved somewhere else.
I've already suggested the tech department of my uni to change to a system which automatically tracks the animals. But that would be too late for me anyway if it ever happens.
>>51997792
<... srsly?
For what purpose? did you feed them some drugs and wanted to see if it altered their behavior?
Isn't it just going to be:
>wake up
>move towards food
>eat
>move to sleeping spot
>fart
>sleep
>GO TO 10
>pink pigs
>what color should be paint the numbers on them anon?
how are you this retarded?
>>51997875
That's a chimpansee. I said I was working with pigs. Since the program I have to use is only available on a laptop provided by my uni, and I am now typing on my own PC, I didn't bother to take a screenshot.
Anyway the research is about the influence of LED (white) lighting systems in comparison to old TL (yellow) lights. Whether the weaned piglets eat more, show more resting behaviour and show less aggressive behaviour in comparison to the pigs housed in the old TL system.
>>51997912
A pink spray can was the only available and provided way of marking the pigs. As the farmer never needs the marking to be on the pigs longer than a day, and it's easy to see with your own eyes it never proved a problem. However it does on video.
I know a lot of this are bad decisions, but we didn't have much influence on the available materials/equipment.
>>51997919
That's pretty cool.
Can't you just choose 2 of the pigs and track those over the 72hours?
>>51997942
Well then you can conclude that LED litening worsens the ability to distinguish pink paint on pigs. And forget about the purpose of the project.
Who knows, that conclusion may one day actually save someone's life instead.
>>51997949
No. Though we are not observing all 72 hours. We're picking different 1 hour timings throughout 3 different periods in the day, ea morning/afternoon/evening.
Each pen houses around 11-12 pigs, which all have to be observed. Since we're going to run statistics over the obtained results we need as much data as possible so to eliminate any statistical errors.
>>51998008
Well then I bid you good luck.
>>51997991
Right on. We actually did another 24 hours of video later on, as to which I got a blue marker for the pigs. Sadly we're not going to use that data, as the client wants to know the difference when the piglets are new to their environment. The first video with pink was taken the first 3 days after the pigs arrived. The second videos were taken 4 weeks in.
