Automatic Coffee Bean Quality Detection System.

This project is the backend of my computer vision-based thesis project. I previously trained a machine learning model that allows the model to distinguish coffee with physical defects such as cracks, holes, black spots, etc., and various types of contaminants that may have entered the coffee bean batch. This Automatic Coffee Bean Quality Detection System is designed for flexible installation on local servers or laptops with adequate GPU support. This system enables efficient AI model inference processing. Its main features are as follows:

• - Real-time Detection API: Provides an application programming interface (API) to detect contaminants and physical defects in coffee beans in real time.
• 
  
• - Live Photo Capture: Supports the ability to instantly capture and process coffee bean photos, for example from a connected camera, for quality analysis.
• 
  
  With this capability, the project's primary goal is to improve the speed, consistency, and accuracy of coffee bean quality assessment, ultimately supporting operational efficiency in the coffee industry.