Bottle Liquid and Cap Detection using OpenCV

This project uses OpenCV (cv2) to analyze images and detect:

1. The amount of liquid in a bottle (as a percentage of the total capacity).
2. Whether the bottle has a cap or not.

This solution is particularly useful for quality control in production lines or automation systems where liquid levels and cap presence need to be verified.

---

Table of Contents

• Features
• How It Works
• Requirements
• Installation
• Usage
• Example Results

---

Features

• Liquid Level Detection: Detects the height of the liquid in the bottle and calculates the percentage of liquid relative to the total bottle height.
• Cap Detection: Identifies whether a cap is present on the bottle.
• Bounding Boxes: Visualizes the detected liquid level, bottle boundaries, and cap presence on the image.

---

How It Works

1. Reference Image: A reference image of a fully filled bottle is used to calculate the height of the cap and the total height of the bottle.
2. Liquid Level Detection:
   • Converts the image to grayscale.
   • Applies Gaussian blur to reduce noise.
   • Uses the Canny edge detection algorithm to find edges.
   • Identifies the contour with the largest area to determine the liquid level.
3. Cap Detection:
   • Crops the top portion of the image based on the liquid height.
   • Processes the cropped region to detect the largest contour.
   • If the contour area is smaller than a predefined threshold, it determines the cap is missing.
4. Output: Displays the processed image with visual indicators (bounding boxes) and prints the results in the console.

---

Requirements

The project requires the following:

• Python 3.8 or newer
• OpenCV

---

Installation

git clone https://github.com/pzimnota/liquid_detection.git

---

Usage

1. Place the images in the Foto directory:

   • Full.jpg: A reference image of a fully filled bottle.
   • Test.jpg: An image of the bottle to be examined.

2. Run the script:

   • python Detection.py

3. Results:

   • The processed image will be saved as Foto/Solution.jpg.
   • The liquid level percentage and cap detection status will be displayed in the console.

---

Example Results

Input:

Reference Image: A fully filled bottle	Test Image: A bottle with varying liquid levels

Output:

• The program overlays bounding boxes:
  Green: Liquid level.
  Blue: Bottle boundaries.
  Red: Cap (if present).

Example:

Bottle without cap	Bottle with cap detected

Console Output:

Bottle without cap	Bottle with cap detected

---