Graph Extractor

Automatically extracts data points from graph images using computer vision, OCR, and signal processing. The extracted data can be analyzed and exported to CSV for further use.

Features

• Detect X and Y axes from graphs
• Extract axis values using OCR (Tesseract)
• Detect grid lines and map pixel positions to actual values
• Digitize graph data points
• Highlight deviation points (local maxima/minima)
• Compute statistics (mean, min, max, RMS, skewness, kurtosis)
• Export data to CSV for analysis

Installation

1. Clone the repository:

   git clone https://github.com/meekhumor/Graph-Extractor.git
   cd Graph-Extractor

2. Install dependencies:

   pip install opencv-python numpy pytesseract scipy pandas

3. Make sure Tesseract OCR is installed and added to your PATH.

Usage

1. Place your graph image in the graph/ directory (e.g., graph/your_graph.png). The script expects a simple line graph with grid lines for best results.
2. Run the main script:

   python extractor.py

   • By default, it processes graph/graph3.png.
   • To use a different image, modify the line:

     preprocess_image('graph/your_image.png')

   • OpenCV windows will display visualizations of axes detection, grid lines, data points, and deviations.
   • Extracted data is saved as csv/graph3.csv (columns: X, Y).
   • Statistics (mean, min, max, RMS, peak-to-valley, skewness, kurtosis) are printed to the console.
3. Press any key in the OpenCV windows to close them.

Example Output

Console Stats:

Mean: 5.2341
Min: 1.0000
Max: 10.0000
RMS: 5.6789
Peak-to-Valley: 9.0000
Skewness: 0.1234
Kurtosis: -0.5678

CSV File (csv/graph3.csv):

X,Y
0.0,4.5
1.0,5.2
2.0,3.8
...

Screenshots

Value Extraction:

Grid Lines:

Data Points:

How It Works

1. Preprocessing: Loads the image, converts to grayscale, and applies thresholding for edge detection.
2. Axis Detection: Uses Hough Line Transform to identify horizontal (X-axis) and vertical (Y-axis) lines.
3. Value Extraction: Crops text regions from axes and uses Tesseract OCR to parse numerical values.
4. Grid Detection: Scans for vertical/horizontal lines, clusters them, and maps to interpolated axis values.
5. Digitization: Finds non-zero pixels (data points), interpolates pixel coordinates to real values.
6. Analysis: Detects local extrema, computes statistics, and exports to CSV.

Directory Structure

meekhumor-graph-extractor/
├── README.md
├── extractor.py
├── graph/          # Input graph images (e.g., graph3.png)
└── csv/            # Output CSV files (auto-created)

Limitations

• Best for simple 2D line plots with visible grid lines.
• OCR accuracy depends on text clarity; may fail on handwritten or stylized labels.
• No support for logarithmic scales or multi-line graphs (extendable).
• Requires manual image path updates; consider adding command-line arguments for production use.

Contributing

Contributions are welcome! Fork the repo, make changes, and submit a pull request. For major changes, please open an issue first.