cyrilli/pose-estimation_python-opencv

Camera calibration and pose estimation with Raspberry Pi and Pi Noir Camera

pose-estimation_python-opencv

Camera calibration and pose estimation of a chessboard using solvePnP Using Opencv-python to estimate pose of a chessboard. Run camcalib.py first to calibrate the camera and get the camera matrix and distortion coefficient. Then run solpnp.py to estimate the extrinsic matrix.

1. PoseEstimateOnRaspberryPi

This folder contains the program I used to run on Raspberry pi. PyQt is used to make the GUI. Run main.py

Required packages include:

1. PyQt4
2. OpenCV3
3. imutils（for the raspberry pi camera）
4. numpy

1.1 The GUI on Raspberry Pi

1.2 Pi NoIR infrared camera

2. MatlabModelForProjection

This folder contains a matlab model of the camera. You can specify the parameters of the intrinsic and extrinsic matrix and watch how the image of the chessboard changes.

2.1 The model of chessboard made up of 3D points

2.2 The projected 2D image with user-specified parameters