Camera Calibration using Zhang’s Method

This project implements Zhang’s method for camera calibration. Camera calibration is essential in computer vision to obtain the intrinsic and extrinsic parameters of a camera, which are used to correct image distortions and understand the 3D geometry of the scene.

Theory

Zhang’s Camera Calibration Method

Zhang’s method is a well-known technique for camera calibration using a planar calibration object, such as a chessboard pattern. The method involves capturing multiple images of the calibration object at different orientations and positions. The main steps are:

1. Image and World Points Extraction:

   • Detect and extract 2D image points (corners of the chessboard pattern) from the calibration images.
   • Generate corresponding 3D world points based on the known dimensions of the chessboard.

2. Homography Computation:

   • Compute homographies between the world points and the image points for each calibration image.

3. Intrinsic Parameters Calculation:

   • Construct a system of linear equations using the homographies to solve for the intrinsic parameters of the camera.

4. Extrinsic Parameters Calculation:

   • Using the intrinsic parameters, compute the extrinsic parameters (rotation and translation) for each image.

5. Distortion Coefficients:

   • Estimate distortion coefficients to account for lens distortion.

6. Optimization:

   • Optimize the intrinsic, extrinsic parameters, and distortion coefficients to minimize the reprojection error.

Prerequisites

To run the code, you need to have the following libraries installed:

• numpy
• opencv-python
• matplotlib
• scipy

You can install these libraries using pip:

pip install numpy opencv-python matplotlib scipy

To run the code (calibrate your camera with custom dataset)

Put all of your images in the “Calibration_Imgs” folder then run the Wrapper file.

python3 Wrapper.py

Results