Calibration of an RTSP stream camera

I am looking at purchaseing MATLAB but have a few questions:

I've seen a lot of information and videos on calibration of cameras but I've not seen the use of that calibration in real time. Once you calibrate the camera(s), can you use the cameras in real time? (I would assume so, but want to confirm)

If you can use the distortion correction in real time, what kind of overhead is used? (I'm trying to determine how much CPU might be used and what is leftover for motion detection and recording.)

Answer 1

[Aastha]

Hi Tboon,

As I understand, you want to know about the computational requirements of using camera calibration parameters, specifically for distortion correction.

Kindly refer to the link of the below example of a camera calibration routine. On running the camera calibration in the example, you obtain the intrinsic, extrinsic, and distortion coefficients for the camera.

https://www.mathworks.com/help/vision/camera-calibration.html

The images captured by the camera often have artifacts that can be corrected using the distortion coefficients to produce undistorted images. This process can be performed using the following MATLAB code:

I = imread('image.png');  
imageSize = [size(I, 1), size(I, 2)]; %calculate the rows and columns 
cameraParams = estimateCameraParameters(imagePoints, worldPoints, 'ImageSize', imageSize);%get camera paramenters in cameraParams object 
J1 = undistortImage(I, cameraParams);%remove distortion 

To get an estimate of CPU usage, information on camera calibration routine and image un-distortion process, you may refer to the link of MathWorks documentation mentioned below:

https://www.mathworks.com/help/vision/ref/undistortimage.html

For precise execution time measurements, you can consider profiling your code to improve performance. For further information on “profiling”, you may refer to the link mentioned below:

https://www.mathworks.com/help/matlab/matlab_prog/profiling-for-improving-performance.html

I hope this helps!