At initial blush, it may well look like tasks that make substantial use of laptop vision or machine learning would require to be dependent on impressive computing platforms with a good deal of clock cycles and memory to cope with this type of software. Although there is some reality to this, as the subject progresses it gets to be probable to experiment with these resources on lower-energy devices as nicely. Get this OpenCV task which is constructed totally on an ESP32 for example.
With that remaining reported, there are some modifications that need to be created to the ESP32 in order to use OpenCV in any significant way. The most crucial of these is the use of the ESP32-DOWDQ6 module which will increase the accessible memory of the ESP32 to allow for it to make greater use of digital camera features. Even then, the ESP32 just can’t run the total OpenCV application, so a shrunken edition of OpenCV is essential before the product can operate it natively. Once these two obstacles are out of the way, although, performing things like edge detection, as this venture demonstrates, are effectively in the realm of possibility.
If managing OpenCV on a thing as tiny as an ESP32 is feasible, it is even less complicated to operate on a thing orders of magnitude additional potent and nevertheless continue to reasonably priced, this sort of as the Raspberry Pi. Even though the project’s code is readily available on its GitHub web site for these fascinated, there are a good deal of other OpenCV assignments that we have highlighted on far more highly effective platforms as well, like this clock which falls off of the wall whenever a person seems to be at it.
Many thanks to [ninjan33r] for the suggestion!