Discuss the working principle behind your project.

Hardware

• Lighting
  • Illumination of the subject of interest (resistor)
  • Using plain backdrop (black/white) for preliminary testing
• Focal length
  • Simple camera with fixed focal length → manual adjustments in distance 
• Adjustments using stepper motor
  • Arduino MEGA
• Image capture
  • Camera module for Raspberry Pi 4
  • Still image vs live footage
  • Limited processing power for live footage (significant lagging)

In order to obtain information about the resistor, we would have to capture it and store the color information, similar to how our eyes process the image we see. We are using an 8-megapixel camera with a fixed focal length and attached a macro lens to it so that it can focus on small objects. Since the camera is of a fixed focal length, we will be installing it on a stepper motor system so that users can fine-tune both the vertical and lateral adjustments.

Software

• OpenCV in Python
• Define color range with Hue/Saturation/Value (HSV) instead of the common Red/Green/Blue (RGB)
  • HSV more resistant to changing light conditions, harder to define the range of colors in RGB
• Finding contour of a resistor using haar cascade classifier
• Filters (to reduce noise and smoothen image): 
  • cv2.medianBlur()
  • cv2.GaussianBlur()
  • cv2.morphologyEx()

The result of our identification of the resistor will be displayed on an LCD screen.