-
Using ESP32-CAM (Ai Thinker) and AI on edge (refer to DroneBot Workshop and EloquentArduino library) to identify different mouse from Logitech.
-
There are 2 optical mouse to identify from the dataset, which are K850 mouse and M221 mouse respectively.The images are captured using ESP32-CAM from different angles and has total of 449 images. Image is resized into 48px height and width, color depth changed to Grayscale .
- K850 mouse
- M221 mouse
esp23cam_webserver_capture is from EloquentArduino library to capture images using Webserver using ESP32-CAM AI Thinker. Object_detection_inferencing folder is downloaded from Edge Impulse to run in Arduino IDE. You can go to the project that I published in Edge Impulse and download the folder as zip.
To use it, simply choose Sketch from the top of Arduino IDE and navigate to Include library and go Add .ZIP Library .... Link to clone my public project in Edge impulse
For using the simplier version with Serial Monitor for showing the predictions, use the EdegeImpulse_FOMO_mouse.ino from the folder , it is the code from EloquentArduino that I modified the line to include the AI model from Edge Impulse instead to use it.