# camera-calibration-with-opencv-python

**Repository Path**: elorfiniel/camera-calibration-with-opencv-python

## Basic Information

- **Project Name**: camera-calibration-with-opencv-python
- **Description**: Calibrate web cameras with Opencv-Python.
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-11-11
- **Last Updated**: 2025-06-23

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Camera Calibration with OpenCV-Python

A simple Python project that demonstrates how to calibrate a camera using OpenCV. Eg. a program takes a set of chessboard images as input, calculates the camera matrix and distortion coefficients, and then uses these parameters to undistort images taken with the same camera.

This project follows the steps outlined in the OpenCV documentation: [OpenCV-Python Tutorials](https://docs.opencv.org/4.10.0/d6/d00/tutorial_py_root.html).

## Note on Dependencies

We use Poetry to manage dependencies (which requires Python 3.8 and above). For further information, see the [Poetry documentation](https://python-poetry.org/docs/). However, the code should run without any issues if you have other versions of Python or other versions of any dependency listed in `pyproject.toml`. Again, this project is for demonstration purposes only.

## Getting Started

Please use the following steps to get started. If you prefer to use another method, eg. `conda`, feel free to do so.

1. Clone this repository with `git clone`, or download the ZIP file.

    ```shell
    git clone <project-repo> <project-path>
    ```

2. Install Pipx, then install Poetry.

    ```shell
    pip install --user pipx
    pipx install poetry
    ```

3. Create a new virtual environment with `venv`.

    ```shell
    python -m venv --upgrade-deps venv

    venv/Scripts/activate.bat # (windows: cmd)
    venv/Scripts/Activate.ps1 # (windows: pwsh)
    source venv/bin/activate  # (linux / mac)
    ```

4. Install the dependencies with `poetry install`.

    ```shell
    cd <project-path> && poetry install
    ```

## Chessboard Pattern

Use `capture.py` to capture chessboard images. Here's the help message for the script:

```text
usage: capture.py [-h] --folder FOLDER [--cap CAP] [--api API] [--w W] [--h H] [--fps FPS] --shape w h name

capture chessboard and save frames

optional arguments:
  -h, --help       show this help message and exit

script options:
  runtime settings used by top-level scripts

  --folder FOLDER  images will be saved in the images folder under this folder

base options:
  calibration name, camera settings and pattern shape

  name             calibration name
  --cap CAP        capture device for cv2.VideoCapture, eg. camera index
  --api API        capture api for cv2.VideoCapture, eg. CAP_ANY
  --w W            capture width, in pixels
  --h H            capture height, in pixels
  --fps FPS        capture framerate
  --shape w h      pattern shape, # of squares, eg. ('w', 'h')
```

## Camera Parameters

Use `calibrate.py` to calibrate the camera parameters. Note that we invoke the `intrinsic` subcommand or the `extrinsic` subcommand accordingly to calibrate the intrinsics or the extrinsics.

```text
usage: calibrate.py [-h] --folder FOLDER [--cap CAP] [--api API] [--w W] [--h H] [--fps FPS] --shape w h {intrinsic,extrinsic} ... name

calibrate camera with chessboard images

optional arguments:
  -h, --help            show this help message and exit

script options:
  runtime settings used by top-level scripts

  --folder FOLDER       images are saved in the images folder under this folder

subcommands:
  calibration type, each has its own options

  {intrinsic,extrinsic}
                        specify type to see additional help
    intrinsic           calibrate camera intrinsics
    extrinsic           calibrate camera extrinsics

base options:
  calibration name, camera settings and pattern shape

  name                  calibration name
  --cap CAP             capture device for cv2.VideoCapture, eg. camera index
  --api API             capture api for cv2.VideoCapture, eg. CAP_ANY
  --w W                 capture width, in pixels
  --h H                 capture height, in pixels
  --fps FPS             capture framerate
  --shape w h           pattern shape, # of squares, eg. ('w', 'h')
```

### Extra Options for `intrinsic` Subcommand

```text
usage: calibrate.py intrinsic [-h] [--wsize WSIZE] [--mm MM]

calibrate camera intrinsics with chessboard images

optional arguments:
  -h, --help     show this help message and exit
  --wsize WSIZE  half of the side length of the search window, see cv2.cornerSubPix
  --mm MM        side length of each square, in millimeters
```

### Extra Options for `extrinsic` Subcommand

```text
usage: calibrate.py extrinsic [-h] --intrinsic INTRINSIC [--wsize WSIZE] [--mm MM]

calibrate camera extrinsics with chessboard images

optional arguments:
  -h, --help            show this help message and exit
  --intrinsic INTRINSIC, --in-params INTRINSIC
                        intrinsic calibration results, in JSON format
  --wsize WSIZE         half of the side length of the search window, see cv2.cornerSubPix
  --mm MM               side length of each square, in millimeters
```