# Object-specific allocator for cv::Mat **URL:** \ **Category:** C++\ **Created:** [September 28, 2022, 10:13pm UTC](https://forum.opencv.org/t/object-specific-allocator-for-cv-mat/10355 "2022-09-28T22:13:38Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![fijoy](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/fijoy/32/882_2.png) [@fijoy](https://forum.opencv.org/u/fijoy)\ **Post date:** [September 28, 2022, 10:13pm UTC](https://forum.opencv.org/t/object-specific-allocator-for-cv-mat/10355/1 "2022-09-28T22:13:38Z")
Does OpenCV support object-specific allocators for cv::Mat class? I’m using OpenCV’s GPU routines and would like to allocate some of the cv::Mat objects using pinned memory. As per [this link](https://stackoverflow.com/questions/62166540/how-to-allocate-opencv-mat-image-on-cuda-pinned-memory), I can call the cv::Mat::setDefaultAllocator() routine to use pinned memory, but that routine is a static member that will make all Mat objects use pinned memory, correct? As an alternative to using cv::Mat::setDefaultAllocator(), I looked into setting the public “allocator” attribute of a Mat object. But as per [this Github issue](https://github.com/opencv/opencv/issues/21637), the allocator attribute is not really used. Is there any other way I can set the allocator for specific Mat objects? ---
**Author:** ![cudawarped](https://avatars.discourse-cdn.com/v4/letter/c/9dc877/32.png) [@cudawarped](https://forum.opencv.org/u/cudawarped)\ **Post date:** [September 29, 2022, 6:06am UTC](https://forum.opencv.org/t/object-specific-allocator-for-cv-mat/10355/2 "2022-09-29T06:06:35Z")
> [@fijoy](#): > > Is there any other way I can set the allocator for specific Mat objects? You could allocate and then pin the memory afterwards using [void cv::cuda::registerPageLocked ( Mat& _m_ )](https://docs.opencv.org/4.x/d9/d41/group __cudacore__ struct.html#ga6d25da8194cc95035994ae98e9eebc02) or just use [cv::cuda::HostMem](https://docs.opencv.org/3.4/d0/d44/classcv_1_1cuda_1_1HostMem.html) directly. ---
**Author:** ![fijoy](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/fijoy/32/882_2.png) [@fijoy](https://forum.opencv.org/u/fijoy)\ **Post date:** [October 3, 2022, 9:52pm UTC](https://forum.opencv.org/t/object-specific-allocator-for-cv-mat/10355/3 "2022-10-03T21:52:01Z")
Thanks - using [cv::cuda::registerPageLocked ( Mat& _m_ )](https://docs.opencv.org/4.x/d9/d41/group __cudacore__ struct.html#ga6d25da8194cc95035994ae98e9eebc02) worked! Using pinned memory improved the performance of my application by 47%. Out of curiosity, I tried cv::Mat::setDefaultAllocator() also, but that had no effect on performance. The exact call I had in my application was: cv::Mat::setDefaultAllocator(cv::cuda::HostMem::getAllocator(cv::cuda::HostMem::AllocType::PAGE\_LOCKED)); Is there an issue with this OpenCV routine, or am I missing something? The OpenCV version I’m using is 4.5.3. ---
**Author:** ![cudawarped](https://avatars.discourse-cdn.com/v4/letter/c/9dc877/32.png) [@cudawarped](https://forum.opencv.org/u/cudawarped)\ **Post date:** [October 4, 2022, 5:59am UTC](https://forum.opencv.org/t/object-specific-allocator-for-cv-mat/10355/4 "2022-10-04T05:59:54Z")
> [@fijoy](#): > > Thanks - using [cv::cuda::registerPageLocked ( Mat& _m_ )](https://docs.opencv.org/4.x/d9/d41/group __cudacore__ struct.html#ga6d25da8194cc95035994ae98e9eebc02) worked! Using pinned memory improved the performance of my application by 47%. That’s great glad I could help! I find `cv::cuda::HostMem` is better if you can use it because you don’t need to remember to call `cv::cuda::unregisterPageLocked()` when your done. > [@fijoy](#): > > Out of curiosity, I tried cv::Mat::setDefaultAllocator() also, but that had no effect on performance. The exact call I had in my application was: > > cv::Mat::setDefaultAllocator(cv::cuda::HostMem::getAllocator(cv::cuda::HostMem::AllocType::PAGE\_LOCKED)); > > Is there an issue with this OpenCV routine, or am I missing something? The OpenCV version I’m using is 4.5.3. I wouldn’t have thought there would be an issue with that version of OpenCV. I just confirmed on the master that changing the allocater forced all alocations of `cv::Mat` to be pinned. If you have built a debug version I would check that this line is hit > when allocating new `cv::Mat`’s after changing the default allocator. ---
**Author:** ![fijoy](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/fijoy/32/882_2.png) [@fijoy](https://forum.opencv.org/u/fijoy)\ **Post date:** [October 4, 2022, 3:37pm UTC](https://forum.opencv.org/t/object-specific-allocator-for-cv-mat/10355/5 "2022-10-04T15:37:38Z")
Thanks! Confirming that cv::Mat::setDefaultAllocator() not working for me was an error on my part. Resolved now. One clarification: pinning with cv::cuda::HostMem is useful only when new cv::Mat objects are allocated, correct? And you have to know the size and type of the allocation also? For example, if I want a pinned cv::Mat after a call to cv::imdecode(), I cannot use cv::cuda::HostMem, unless I know in advance the size and type of the image being decoded? ---
**Author:** ![cudawarped](https://avatars.discourse-cdn.com/v4/letter/c/9dc877/32.png) [@cudawarped](https://forum.opencv.org/u/cudawarped)\ **Post date:** [October 4, 2022, 4:09pm UTC](https://forum.opencv.org/t/object-specific-allocator-for-cv-mat/10355/6 "2022-10-04T16:09:45Z")
I would guess from the function signature ```auto cv::imdecode (InputArray buf, int flags, Mat *dst) ``` that you can’t pass a `HostMem` object as `dst` because it is a `Mat` and not an `OutputArray`. That said I know `VideoCapture::read` works and I would expect all functions which the output is an `OutputArray` to allow you to pass in an empty `HostMem` object and internally alocate it for you.