Raspberry Pi & Tensor Flow based project

Thread Starter

Arjunreddy

Joined Jul 14, 2021
3
I am working on a project involving Raspberry pi & tensor flow to count lighted LEDs and generate analog signals accordingly. Any ideas or projects related or references.
PI 3 with camera, Python
Thank you.
 

FlyingDutch

Joined Mar 16, 2021
83
Hello,

you have to describe much more precisely the algorithm in order to somebody be able help you. Describe algorithm of generation analog signals depending on LEDs state. Maybe AI (ANN) and Tensorflow not have to be needed, and just library for computer vision (for example OpenCV) be sufficient.

Best Regards
 

Thread Starter

Arjunreddy

Joined Jul 14, 2021
3
Hello,

you have to describe much more precisely the algorithm in order to somebody be able help you. Describe algorithm of generation analog signals depending on LEDs state. Maybe AI (ANN) and Tensorflow not have to be needed, and just library for computer vision (for example OpenCV) be sufficient.

Best Regards
The LEDs are in series parallel pattern, so can't really use current sensor.
When all LEDs are ON, one of gpio should go high, so that I can use that signal to guide a robot to put the board to a different location and if minimum one is OFF, Some other gpio should go high to place the board in some other location.
 

FlyingDutch

Joined Mar 16, 2021
83
Hello,

for such algorithm you don't need necessarily ANNs and Tensorflow (easier way should be enough) , just OpenCV library itself should be enough. See for example these links:

https://www.pyimagesearch.com/2016/...ght-spots-in-an-image-with-python-and-opencv/

https://medium.com/analytics-vidhya/detecting-and-counting-objects-with-opencv-b0f59bc1e111

https://www.hackster.io/phfbertoleti/counting-objects-in-movement-using-raspberry-pi-opencv-015ba5

https://answers.opencv.org/question/206921/object-detection-and-counting/

Use google to find more examples of your subject.

Best Regards
 
Top