3d_ImgProcessing

This repository consists of Matlab scripts to demonstrate the image processing starting from the very basics of Image-data types, color-models to calibration of mono and stereo cameras and setting up the multiple camera models for the Husky robot and further processing to SIFT detector and discriptor for image stitching and feature matching.

Img_datatypes:

This package consists of Matlab scripts to understand the different Image types and conversion of images from one type to the other using standard Matlab libraries

• Interconversion of GrayScale img, RGB img, Binary img and image indexing
• Thresholding using Otsu principle and histogram creation
• Introduction to indexing of the image and colormap

Img_colormodels:

This package consists of Matlab scripts to understand the different color spaces such as RGB, HSV, LAB, YIQ, YUV, XYZ and GrayScale. Conversion of images from onr colorspace to the other and analysis of the image data mathematically using Matlab standard libraries.

• Extracting the Red, Blue, and Green channels from RGB image and generating the corresponding color map
• Creating grayscale images with linear combination of different components and averaging methods
• Conversion of RGB image to HSV colorspace, extracting the Hue, Saturation, and Value channels from HSV image, thresholding the individual channels and reconstructing the image

Mon_cam_calibration:

This package consists of Matlab scripts to capture the images and perform the camera calibration using Matlab.

• Start the USB camera, and record the checkerboard images for camera calibration

• Detect the corner points of the checkerboard and calculate image_points and board size parameters

• Transform the image points from the corner points into world coordinates

• Calculate the intrinsic and extrinsic camera parameters

• Visualize the reprojection error and extrinsic parameters of the calibrated camera

  

Stereo_cam_calibration:

This package consists of Matlab scripts to capture the images and perform the stereo camera calibration using Matlab.

• Start the USB camera, and record the checkerboard images for camera calibration

• Detect the corner points of the checkerboard and calculate image_points and board size parameters

• Transform the image points from the corner points into world coordinates

• Calculate the intrinsic and extrinsic camera parameters from the stereo camera pair

• Visualize the reprojection error and extrinsic parameters of the calibrated stereo camera system

• Rectification of the stereo image pair

• Compute the disparity between stereo image pairs and reconstruction of the 3D scene using a disparity map and reprojection matrix

  

Multi_cam_model:

This package consists of Matlab scripts to capture the images perform the stereo camera calibration of two stereo pairs and visualize the extrensics of all cameras with respect to the checkere board using Matlab.

• Stereo camera calibration of the two stereo pairs, then visualize the camera extrinsic in the same coordinate system using the rotation and translation vectors obtained from the camera calibration process w.r.t checkered board co-ordinate system.