This post shows the benefit of using v3dfx-base in a complex end-user application - An Ultrasound Image Viewer.
This Qt application loads a real Ultrasound file consisting of 128 layers of captured images, renders using v3dfx-base in a QGraphicsView using V3dfxGLScene, blends all 128 layers of them to recreate the final image. With a regular glTexImage2D approach, this needs 100% CPU loading. With the v3dfx-base, this needs < 8% CPU loading on SGX530@200MHz (Beaglebone/3730 class). This show-cases the powerful capabilities of the SGX with image/video streaming use-cases in both ARGB and YUV modes.
Screenshot:
https://github.com/prabindh/v3dfx-base/blob/master/test/platforms/qt_ultrasound/Photo0039.jpg
README, and complete code:
https://github.com/prabindh/v3dfx-base/tree/master/test/platforms/qt_ultrasound
Please see earlier blog post for introduction to v3dfx-base.
No comments:
Post a Comment