Liquid Level Detection in Bottles 🧪📏

This project presents five different approaches to detect liquid levels in bottles using image processing techniques with Python and OpenCV.

🔍 Methods Overview

The repository includes five separate scripts, each employing a distinct strategy to locate the horizontal liquid surface inside transparent or semi-transparent bottles.

1. color_diff.py – Pixel Intensity Analysis

• Converts the image to grayscale.
• Binarizes and normalizes the image.
• Calculates row-wise pixel intensity (Y-axis).
• Visualizes the intensity distribution as a bar plot.

📊 Good for: Even lighting and clear air-liquid contrast.

---

2. line_overlay.py – Sobel + Morphological Filters

• Applies Sobel Y edge detection to enhance horizontal edges.
• Uses morphological closing and opening to refine those edges.
• Applies Hough Line Transform to detect straight horizontal lines.

🧼 Good for: Sharp edges and images with minimal noise.

---

3. line_detection_draw.py – Hybrid Analysis

• Combines Sobel, morphology, and row-wise intensity summation.
• Outputs both processed images and bar plots of pixel intensity.
• Includes intermediate visualizations for debugging and evaluation.

🧪 Good for: Complex scenes or noisy backgrounds.

---

4. y_coordinate.py – Horizontal Line Clustering

• Extracts horizontal lines using Sobel + Canny + Hough.
• Gathers the Y-coordinates of these lines.
• Applies KMeans clustering to group and highlight representative liquid levels.

📍 Good for: Precise line-level Y-coordinate localization.

---

5. a_b_params.py – Line Equation Clustering

• Detects horizontal-ish lines and converts them to slope-intercept (a, b) form.
• Clusters these lines using KMeans to group dominant line patterns.

📐 Good for: Mathematical analysis and grouping of similar line orientations.

---

📁 Folder Structure

liquid-level-detection/
│
├── a_b_params.py # Slope-intercept (a, b) clustering of horizontal lines
├── y_coordinate.py # KMeans clustering of horizontal Y-coordinates
├── line_overlay.py # Sobel + Morphology + Hough Line method
├── color_diff.py # Grayscale pixel intensity difference method
├── line_detection_draw.py # Combined hybrid method
│
├── requirements.txt # Python dependencies
├── README.md # Project documentation
│
├── img/ # Input images folder
│ └── (your .jpg/.png files here)
│
├── output_folder/ # Output from line_overlay, y_coordinate, a_b_params
│ └── (processed images/results)
│
└── color_diff/ # Output from color_diff.py
├── row_sums_X.jpeg
└── row_sums_X.txt