Home/ Questions/Q 5972649
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-22T20:41:17+00:00

Is there any papers that describe the algorithm of the FindCornerSubPix function in openCV?

  • 0

Is there any papers that describe the algorithm of the FindCornerSubPix function in openCV? I cannot find any documentation which describes it.

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    I believe from the source code (for version 2.1) it is using the image gradients to optimise location…

    cv.h:
997 /* Adjust corner position using some sort of gradient search */
998 CVAPI(void)  cvFindCornerSubPix( const CvArr* image, CvPoint2D32f* corners,
999                                  int count, CvSize win, CvSize zero_zone,
1000                                  CvTermCriteria  criteria );

    The documentation for the function also describe this – particular the nature of the objective function used:

    http://opencv.willowgarage.com/documentation/c/imgproc_feature_detection.html?highlight=findcornersubpix#cvFindCornerSubPix

    More explicit details of the operation are found in the file cvcornersubpix.cpp. Looking at this it can be clearly seen that the function is calculating the derivatives of the search area and building a Hessian matrix…

    In other words, as I understand it, a standard Conjugate gradient method!

    Version 2.2 of OpenCV (the most up-to-date) does not appear to have this function anymore.

    • 0