I have a project that involves connecting 3 ultrasonic distance sensors, 2 motors, 3 cameras and a 9g servo. I'm still very new to this and I don't know how to connect them all together without overloading the Arduino. I'm planning on using the Uno rev 3.
Try to draw and show a proposed schematic for the project.
When said about "overloading the Arduino", did you mean the power problems?
Please explain your project in more detail. Especially this is interesting:
What the cameras do you want to connect? Arduino Uno is hardly will able to interact with one camera, so three is a definitely too much for it.
Its called a rescue maze robot, its for a competition. Basically, it is a 4 wheeled vehicle that moves 30 cm at a time (the size of the tile) and scans left and right for victims. It drops a care package when it sees one. The build I have in mind is 1 sensor each side for alignment and 1 in front for avoiding hitting the walls, 2 cameras on the sides for victim recognition and a camera in front to avoid black tiles and the servo to drop the packages. I am given 2 months to make this project possible with zero experience so your help would be very much appreciated. By the way I haven't bought the arduino yet so if u think a esp32 would be better I can change my mind.
Recording still images from a single camera is only just within the capacity of Uno. 3 cameras might be possible, provided images are taken in turn from each camera.
But you seem to be talking about image recognition, in other words, detecting certain objects within the images, which could be in any orientation and at varying distances from the camera. This is way, way beyond the capabilities of Uno.
I have heard that esp32 is capable of facial recognition. This surprises me and I wonder just how good it could actually be.
Generally speaking, image recognition involves using machine learning algorithms that require significant resources, meaning many fast and powerful CPU cores and very large amounts of memory. No Arduino has those things.
Is there some other, much simpler way your robot can detect these "victims"? You will need that to have any chance of success with zero experience and only 2 months.
1 Like
This is what Im aiming to make. I asked them and they told me they used the esp32 with 8 lidar sensors and 2 or 3 cameras and the motors and the servo of course. you think u could help me figure out how?
I will assume that the robot must complete all these tasks autonomously. You will probably need an ESP32. An Uno will go nowhere in this competition. 3 cameras though? Are there any significant size requirements? Because if you put a camera on a swivel, then you can accomplish what may take 3 cameras with only one camera and an additional servo.
As far as I see, the only decision making the robot will have to do is the care package. Most of the processing power will be dedicated to the camera and image recognition. If you time how much the robot moves in one second, you can use that to turn on and off the motors, which can then move exactly one tile. And, if your motors have encoders, it's even better, because using the encoder ticks allows for far more accuracy. This would also eliminate the need for sensors, although you still may want to include some if the ESP still has available pins (probably does, I mean, 3 cameras and 8 LIDAR sensors, it probably has enough pins).
Ridiculous.
Do some Internet research on image-recognition and take note on the software and the computer board (processor) being used.
robot with multiple cameras and image recognition at DuckDuckGo
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.