Cv::cuda::resize still available in opencv version 4.x?

yuchenw · February 17, 2025, 11:50am

I’m using cv::cuda::resize which is provided by cudawarping hpp. I can’t find the implementation (definition of this function) in version 4.x source code but in version 3.4.
I’m curious that whether in 4.x, the function definition is removed but its implementation is still in dll of version 4.x, correct?

cudawarped · February 17, 2025, 12:39pm

In OpenCV 4.x the CUDA modules were moved to the contrib repo, see

github.com/opencv/opencv_contrib

modules/cudawarping/include/opencv2/cudawarping.hpp

ce3c6681c


      
          @param fx Scale factor along the horizontal axis. If it is zero, it is computed as:
          \f[\texttt{(double)dsize.width/src.cols}\f]
          @param fy Scale factor along the vertical axis. If it is zero, it is computed as:
          \f[\texttt{(double)dsize.height/src.rows}\f]
          @param interpolation Interpolation method. INTER_NEAREST , INTER_LINEAR and INTER_CUBIC are
          supported for now.
          @param stream Stream for the asynchronous version.
          
          @sa resize
           */
          CV_EXPORTS_W void resize(InputArray src, OutputArray dst, Size dsize, double fx=0, double fy=0, int interpolation = INTER_LINEAR, Stream& stream = Stream::Null());
          
          /** @brief Applies an affine transformation to an image.
          
          @param src Source image. CV_8U , CV_16U , CV_32S , or CV_32F depth and 1, 3, or 4 channels are
          supported.
          @param dst Destination image with the same type as src . The size is dsize .
          @param M *2x3* Mat or UMat transformation matrix.
          @param dsize Size of the destination image.
          @param flags Combination of interpolation methods (see resize) and the optional flag
          WARP_INVERSE_MAP specifying that M is an inverse transformation ( dst=\>src ). Only

yuchenw · February 18, 2025, 1:51am

Got it, thanks for your answer, that’s really helpful!
I encountered an issue before which is very similar to this one CUDA GoodFeaturesToTrackDetector is not ThreadSafe ? · Issue #18051 · opencv/opencv · GitHub
When calling cv::cuda::resize() of version 3.4, it crashed because that function uses texture reference, which is quite obsolete and does not support multi-threaded programming.
Now I’m clear that it has been fixed in 4.x because texture feature is no longer being used.

Topic		Replies	Views
How to resize cv::mat image size with a cuda kernel? C++ cuda , imgproc	2	1992	May 27, 2022
Looking for CUDA functions Python java , cuda	8	993	August 3, 2023
GPU Affine Warp C++ cuda , imgproc	2	967	February 4, 2022
Any OpenCV Cuda function have similar used for cv::distanceTransform? C++ cuda , imgproc	4	392	August 14, 2023
Contributing using CUDA - versions and libraries C++ cuda	4	372	December 25, 2023

Cv::cuda::resize still available in opencv version 4.x?

Related topics