OpenCV
OpenCV 4.5.0 for Windows (Tag 4.5.0: source main, source contrib, 12/10/2020).
Note: The CUDA redistributable dll’s are not included in the OpenCV 4.5.0 downloads below. To use these builds you will either have to install both CUDA 11.1 and cuDNN 8.0.4 on your machine or get hold of the redistributable dll’s from an install on another machine.
- VS2019
- OpenCV 4.5.0 x64, VS2019 with CUDA all modules + python 3.8 bindings (Release, Debug ) configured with:
- CUDA 11.1 binaries compatible with compute 3.5-8.6, –-use_fast_math enabled, cv::cuda).
- cuDNN 8.0.4 (DNN_BACKEND_CUDA).
- Nvidia’s NVDECODE hardware video decoder (cv::cudacodec::VideoReader).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.5.0 x64, VS2019 with CUDA DNN backend only + python 3.8 bindings (Release, Debug ) configured with:
- CUDA 11.1 for CUDA DNN backend (binaries compatible with compute 6.0-8.6, cv::cuda).
- cuDNN 8.0.4 (DNN_BACKEND_CUDA).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.5.0 x64, VS2019 with CUDA all modules + python 3.8 bindings (Release, Debug ) configured with:
OpenCV 4.4.0 for Windows (Tag 4.4.0: source main, source contrib, 18/07/2020).
Note: The CUDA redistributable dll’s are not included in the OpenCV 4.4.0 downloads below. To use these builds you will either have to install both CUDA 11.0 and cuDNN 8.0 on your machine or get hold of the redistributable dll’s from an install on another machine.
- VS2019
- OpenCV 4.4.0 x64, VS2019 with CUDA all modules + python 3.7 bindings (Release, Debug ) configured with:
- CUDA 11.0 binaries compatible with compute 3.5-8.0, –-use_fast_math enabled, cv::cuda).
- cuDNN 8.0.1 RC2 (DNN_BACKEND_CUDA).
- Nvidia’s NVDECODE hardware video decoder (cv::cudacodec::VideoReader).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.4.0 x64, VS2019 with CUDA DNN backend only + python 3.7 bindings (Release, Debug ) configured with:
- CUDA 11.0 for CUDA DNN backend (binaries compatible with compute 6.0-8.0, cv::cuda).
- cuDNN 8.0.1 RC2 (DNN_BACKEND_CUDA).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.4.0 x64, VS2019 with CUDA all modules + python 3.7 bindings (Release, Debug ) configured with:
OpenCV 4.3.0 for Windows (Tag 4.3.0: source main, source contrib, 06/04/2020).
Note: The CUDA redistributable dll’s are not included in the OpenCV 4.3.0 downloads below. To use these builds you will either have to install both CUDA 10.2 and cuDNN on your machine or get hold of the redistributable dll’s from an install on another machine.
- VS2019
- OpenCV 4.3.0 x64, VS2019 with CUDA all modules + python 3.7 bindings (Release, Debug ) configured with:
- CUDA 10.2 binaries compatible with compute 3.0-7.5, –-use_fast_math enabled, cv::cuda).
- cuDNN 7.6.5.32 (DNN_BACKEND_CUDA).
- Nvidia’s NVDECODE hardware video decoder (cv::cudacodec::VideoReader).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.3.0 x64, VS2019 with CUDA DNN backend only + python 3.7 bindings (Release, Debug) configured with:
- CUDA 10.2 for CUDA DNN backend (binaries compatible with compute 6.0-7.5, cv::cuda).
- cuDNN 7.6.5.32 (DNN_BACKEND_CUDA).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.3.0 x64, VS2019 with CUDA all modules + python 3.7 bindings (Release, Debug ) configured with:
OpenCV 4.2.0 for Windows (Tag 4.2.0: source, 20/12/2019).
Note: The CUDA redistributable dll’s are not included in the OpenCV 4.2.0 downloads below. To use these builds you will either have to install both CUDA 10.2 and cuDNN on your machine or get hold of the redistributable dll’s from an install on another machine.
- VS2019
- OpenCV 4.2.0 x64, VS2019 with CUDA all modules + Anaconda python 3.7 bindings, configured with:
- CUDA 10.2 binaries compatible with compute 5.3-7.5, –-use_fast_math enabled, cv::cuda).
- cuDNN 7.6.5.32 (DNN_BACKEND_CUDA).
- Nvidia’s NVDECODE hardware video decoder (cv::cudacodec::VideoReader).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.2.0 x64, VS2019 with CUDA DNN backend only + Anaconda python 3.7 bindings, configured with:
- CUDA 10.2 for CUDA DNN backend (binaries compatible with compute 5.3-7.5, cv::cuda).
- cuDNN 7.6.5.32 (DNN_BACKEND_CUDA).
- Intel Quick Sync hardware video encoder/decoder (cv::CAP_INTEL_MFX).
- OpenCV 4.2.0 x64, VS2019 with CUDA all modules + Anaconda python 3.7 bindings, configured with:
OpenCV 4.1.0 for Windows (Tag 4.1.0: source, 08/04/2019).
- VS2017
- OpenCV 4.1.0 x64, VS2017 with CUDA 10.1 + python bindings for CUDA, configured with:
- CUDA 10.1 (binaries compatible with compute 3.0-7.5, –-use_fast_math enabled).
- OpenCV 4.1.0 x64, VS2017 with CUDA 10.1 + python bindings for CUDA, configured with:
OpenCV 4.0.0 for Windows (Tag 4.0.0: source, 18/11/2018).
Note: The CUDA and TBB redistributable dll’s are not included in the OpenCV 4.0 downloads below. To use these builds you will either have to install both CUDA 10.0 and Intel TBB 2018 on your machine or get hold of the redistributable dll’s from an install on another machine.
- VS2017
- OpenCV 4.0 x64, VS2017 with CUDA 10.0, MKL(TBB) and TBB, configured with:
- CUDA 10.0 (binaries compatible with compute 3.0-7.5, –-use_fast_math enabled).
- Intel MKL (2019.1.144) & TBB (2019.2.144).
- OpenCV 4.0 x64, VS2017 with CUDA 10.0 and MKL(TBB), configured with:
- CUDA 10.0 (binaries compatible with compute 3.0-7.5, –-use_fast_math enabled).
- Intel MKL (2019.1.144) & TBB (2019.2.144).
- OpenCV 4.0 x64, VS2017 with CUDA 10.0 and python bindings, configured with:
- CUDA 10.0 (binaries compatible with compute 3.0-7.5, –-use_fast_math enabled).
- OpenCV 4.0 x64, VS2017 with CUDA 10.0, MKL(TBB) and TBB, configured with:
OpenCV 3.4 for Windows (commit: 6d4f66472e14b29b8e1623859cfebfdc67f677c3, 22/12/2017).
Note: The CUDA and TBB redistributable dll’s are not included in the OpenCV 3.4 downloads below. To use these builds you will either have to install both CUDA 9.1 and Intel TBB 2018 on your machine or get hold of the redistributable dll’s from an install on another machine.
- VS2017
- OpenCV 3.4 x64, VS2017 with CUDA 9.1, MKL(TBB) and TBB, configured with:
- CUDA 9.1 (binaries compatible with compute 3.0-7.0, –-use_fast_math enabled).
- Intel MKL (2018.0.124) & TBB (2018.0.124).
- OpenCV 3.4 x64, VS2017 with CUDA 9.1 and MKL(TBB), configured with:
- CUDA 9.1 (binaries compatible with compute 3.0-7.0, –-use_fast_math enabled).
- Intel MKL (2018.0.124) & TBB (2018.0.124).
- OpenCV 3.4 x64, VS2017 with CUDA 9.1, configured with:
- CUDA 9.1 (binaries compatible with compute 3.0-7.0, –-use_fast_math enabled).
- OpenCV 3.4 x64, VS2017 with CUDA 9.1, MKL(TBB) and TBB, configured with:
OpenCV v3.3 for Windows (commit: 4af3ca4e4d7be246a49d751a79c6392e848ac2aa, 04/08/2017).
- VS2015
- OpenCV v3.3 x64, VS2015 with CUDA
Decryption key: !PgY6i9jLWD4398lLHpy4DdNYATVnbfsnmpmNsy1ptxc
Configuration:- CUDA 8.0.61 + Patch 2 (binaries compatible with compute 2.0-6.1, –-use_fast_math enabled).
- Intel MKL (2017 update 3) & TBB (2017 update 6).
- OpenCV v3.3 x64, VS2015 with CUDA
- VS2013
Note: The CUDA and TBB redistributable dll’s are not included in the download below. To use these builds you will either have to install both CUDA 8.0 and Intel TBB 2018 on your machine or get hold of the redistributable dll’s from an install on another machine.- OpenCV 3.3 x64, VS2013 with CUDA, MKL(TBB), TBB and python bindings
Configuration:- CUDA 8.0.61 + Patch 2 (binaries compatible with compute 2.0-6.1, –-use_fast_math enabled).
- Intel MKL (2018) & TBB (2018).
- Bindings for both Anaconda Python 2 and 3.
- OpenCV 3.3 x64, VS2013 with CUDA, MKL(TBB), TBB and python bindings
OpenCV v3.2 for Windows (commit: 70bbf17b133496bd7d54d034b0f94bd869e0e810, 23/12/2016).
- VS2015
- OpenCV v3.2 x64, VS2015 with CUDA
Decryption key: !gxqpnR0Qol9NZbfAjxZvjFG-iXuxsDpsvxQGK72CK3c
Configuration:- CUDA 8.0 (binaries compatible with compute 2.0-6.1, –-use_fast_math enabled).
- Intel MKL (2017 update 1) & TBB (2017 update 2).
- Eigen 3.33.
- SIMD optimizations (SSE, SSE2, SSE3, SSE41, SSE42, SSSE3, AVX, AVX2, FMA3).
- OpenCV v3.2 x86, VS2015 without CUDA
Decryption key: !tIZh7bPHLq1zm_mQkUDAarJELe6BNSxN8cvtvl3FYyo
Configuration:
- OpenCV v3.2 x64, VS2015 with CUDA
- VS2013
- OpenCV v3.2 x64, VS2013 with CUDA
Decryption key: !OSYb1OoWm98IYFb-xjOiLjoU6zZbOaG-go3uWpnA1ds
Configuration:- CUDA 8.0 (binaries compatible with compute 2.0-6.1, –-use_fast_math enabled).
- Intel MKL (2017 update 1) & TBB (2017 update 2).
- Eigen 3.31.
- SIMD optimizations (SSE, SSE2, SSE3, SSE41, SSE42, SSSE3, AVX, AVX2, FMA3).
- OpenCV v3.2 x86, VS2013 without CUDA
Decryption key: !U1xhT5dS-Zi42SXajQE2o9wsQQjm5fii3oMeoAqNpXQ
Configuration:
- OpenCV v3.2 x64, VS2013 with CUDA
Copyright secured by Digiprove © 2020 James Bowley