Group Actions
Engagement
  • Author Author: dubbie
  • Date Created: Date Created
  • Views 2200 views
  • Likes 9 likes
  • Comments 7 comments
Related
Recommended

CatDogFoxBot #2 : System Design and Gathering the Components

dubbie

Now that the Remote Monitoring and Control Project14 Challenge has started I have gathered together some of the components and PCBs that I am thinking of using, following on from my initial ideas (CatDogFoxBot #1 : My Initial Idea). The Grideye module was ordered and has arrived and is much smaller than I was expecting, although this is not a problem. I will also be using one of the step motors I used in my HoloPiBot Picasso Design Challenge ( HoloPiBot #10 : It Actually Works - It Really Does!  ). They are not particularly powerful but it should be OK for this small sensor, plus I already have one and they come with their own driver PCB. Having created a nice 3D printer holder for my DC motor mobile robot ( DCMotorsMobileRoBot #4 : 3D Printed PCB Holder  ) I have decided to try and do the same for this project and I will be using another connectorless Nano.

 

 

One of the problems that I anticipate is actually testing out the system to detect cats. We do not get any dogs running about wild in this Country any more and only the occasional fox, but you never know - they might be out there! So I expect to have to run the system for long periods while collecting image data to confirm that it is working and I am thinking of a Bluetooth link to my laptop indoors, saving all the Grideye data into a text file. I could use a WiFi link but I am keeping the new technology as limited as possible and just sticking to the Grideye which I have not used before. It it all goes well and there is still some time left then I might consider trying out a WiFi link as well. To make all this work all night means some sort of big battery and I am going to use a lead acid battery that I just happen to have lying around. This is from a fire alarm battery backup which it is no longer good enough for but should hopefully be good enough for me.

 

If this all works then I will have to think of some sort of environmental packaging and protection and I thought I might try making a concrete or cement enclose for the battery part at least. I have been involved in making one before when testing how good WiFi was at passing through concrete but it turned into a monstrosity that was virtually impossible to lift. I would want to go for something more delicate and smaller for this project.

 

Well, I'm off to design some 3D part now to connect the scanning bits all together.

 

Dubbie

Parents

  • Dubbie,

     

    Looks like things are coming together.  As I was reading/watching, I was thinking about trying to correlate what the grideye is seeing to an actual species.  It might be nice if you had a camera that could capture an image when a possible detection is made.  Then you could possible determine the species and train the system to help log the activity.

     

    Seem a bit more complex, but an interesting problem.

     

    Good luck on your continued progress.

     

    Gene

  • in reply to genebren

    Gene,

     

    I hadn't thought about adding a proper camera as well as my main focus is on the night hours when it is dark, but it is a possibility. I am hoping to be able to recognise whether it is a cat or not just using the Grideye data. There isn't much data to work with but it might be possible to do something, which is why I want to capture all the data, so I can see if it might be possible to process this 'image data to identify cats. It might just end up detecting hot bodies.

     

    Dubbie

  • in reply to dubbie

    Dubbie,

     

    I understand that.  I have used an infrared game camera, which captures a short video based on a PIR motion sense trigger (to determine who was digging up our garden - turned out to be raccoons and skunks, although I was hoping it was an armadillo).  You could run something like this in parallel to your system to help your learn how to interpret your data during development.

     

    Gene

  • in reply to genebren

    Gene,

     

    I do have a couple of cameras (Raspberry Pi V2 camera and a Pixy camera) both of which will work if infrared light is supplied from some LEDs. It is a possibility. I doubt I'll see any armadillos in our garden, or racoons or skunks - maybe a hedgehog if I'm lucky, rats if I'm unlucky!

     

    For the moment I'm going to concentrate on the Grideye data and do some simple image processing as I did some in the past with black and white images and it wold be fun to re-visit those ideas. Plus, I have the whole environmental thing to consider as I do not think I have never made any electronics that was designed to be outside all the time. I have run the odd mobile robot outside but usually only on dry days in daylight. There will be damp, insects, rain, low temperatures, high temperatures (we had over 34 C last week and the plastic tops on the propagators in the greenhouse melted. I've never seen that before.) and probably some other things I haven't even thought of.

     

    Dubbie

  • in reply to dubbie

    Looks like OpenCV has cat face detection built-in.

     

    Detecting cats in images with OpenCV

    https://www.pyimagesearch.com/2016/06/20/detecting-cats-in-images-with-opencv/

     

    Moggie shot ? "Guilty as charged puss" image

Comment Children
  • in reply to beacon_dave

    The first picture made me laugh. It seems more like something from April 1st than an actual thing, but on reflection, there is no reason why cats faces should not be as detectable as human faces, or dogs, or horses, or probably anything. I don't think I will implement this as I don't have the time, knowledge or cat enthusiasm. I'm just going to stick with detecting hot bodies (hopefully).

     

    Dubbie