How to remove distortions from an image which is imported in MATLAB as a matrix of pixel values (as a file.MAT)?

Hi, I'm going to show my problem. By using a thermographic camera an image is recorded and saved as a file.MAT, so in MATLAB I get a matrix which have the same dimensions of the ones of the thermal image. I can print this image by using "imagesc". The problem is that the thermal camera is located with a certain lateral angle with respect the target object which is recorded, so the printed image (with imagesc) appears distorded: it should be be a square image, instead it's a trapeze. My purpose is to remove this distortion. How can I go on? Should I use an interpolation?Furthermore each pixel value in the matrix must be remain the same. Thanks

Answer 1

[Image Analyst]

There is a camera calibration capability in the Computer Vision Toolbox. Check it out.

Also see Steve's blog: Spatial transformations Defining and applying custom transforms Steve on Image Processing

Answer 2

[DGM]

Alternatively, if you don't have CVT, and you're not processing many images, you could use fitgeotrans() and imwarp().

This is one simple example.

https://www.mathworks.com/matlabcentral/answers/456973-how-to-make-an-image-straightener-like-that-of-cam-scanner-app