OpenCV is an open source computer vision library that contains image processing functions that can be used for comparison with reference images, extraction of license plates and so on, and is used in a wide range of fields such as smart factory and smart infrastructure. OpenCV Accelerator utilizes Renesas unique hardware DRP suitable for image processing as an image processing accelerator, and it improves user system performance by accelerating OpenCV image processing.


  • Easy to use
    • OpenCV compatible API allows you to develop your application software without hardware knowledge.
  • Performance
    • Get better performance by using dedicated hardware accelerator for image processing.
  • CPU off-load
    • Reduce CPU load by processing image with dedicated hardware accelerator.


  • Factory automation
  • Smart infrastructure
  • Robotics


OpenCV Accelerator performs part of the function provided by OpenCV with image processing accelerator DRP automatically, allowing you to accelerate image processing and reduce CPU load. The interface is compatible with OpenCV, so hardware knowledge is not needed. Open source application software and existing application software that uses OpenCV can be used without any changes and can be accelerated.

OpenCV Accelerator: Possible to develop without hardware knowledge. Open source and existing application software can be accelerated without any changes.

Functions and Capabilities

OpenCV Accelerator functions provided for RZ/V2MA and their main capabilities

Function Description
cv::resize Resize an image
cv::cvtColor (YUV2RGB) Change color space YUV to RGB
cv::cvtColorTwoPlane (YUV2RGB_NV12) Change color space NV to RGB
cv::GaussianBlur Gaussian filter
cv::dilate Dilates an image
cv::erode Erodes an image
cv::morphologyEx Morphological transformations
cv::filter2D Convolves an image
cv::sobel Sobel filter
cv::adaptiveThreshold Adaptive threshold
cv::matchTemplate Compares a template against overlapped image regions
cv::warpAffine Affine transformation
cv::warpPerspective Perspective transformation
cv::pyrDown Blurs an image and downsamples
cv::pyrUp Upsamples an image and then blurs
OpenCV Accelerator and CPU performance

Performance measurement conditions:

  • RZ/V2MA CPU (Cortex®-A53 Dual 996MHz)
  • Input image resolution is FHD (1920x1080). Processing time depends on setting parameters.

Example of OpenCV Usage

The following example is of appearance inspection in a production line by extracting and checking printed text by image processing using OpenCV. In the following example, Binarization Adaptive and Dilate are performed with OpenCV Accelerator, and Cropping, Find Contours and checking printed text are performed with the CPU. By using OpenCV Accelerator, part of image processing can be accelerated, so the total processing time can shortened. Furthermore, it is possible to achieve better performance through parallel processing of OpenCV Accelerator and the CPU.

Example of OpenCV usage

Software List

RZ/V2MA OpenCV Accelerator Support Package [V1.30]
This product provides OpenCV Accelerator Support Package for RZ/V2MA.
Software Package Renesas
RZ/V2L OpenCV Accelerator Support Package [V1.10]
This product provides OpenCV Accelerator Support Package for RZ/V2L.
Software Package Renesas
2 items