Object Classification and Counting Using Raspberry Pi for Small Items

The Object Classification and Counting system operates by capturing visual data of small items, such as buttons or chocolates, through a connected Camera Module linked to the Raspberry Pi 4 Model B. The Raspberry Pi acts as the central processing unit running a custom machine learning model trained via Edge Impulse. When items are placed within the camera's field of view, the system processes the live video frames locally at the edge, requiring no constant cloud connection. The trained neural network analyzes the pixel data to identify the specific class of each item and increments an internal digital counter accordingly. A Python script manages the logic for capturing images, executing the machine learning inference, and keeping track of the counts in real time. Because the processing happens entirely on the Raspberry Pi 4, the system can instantly output the categorization and total tally, making it a self-contained edge intelligence solution.