Hello,everyone
I am using a DM8168 board with DVRRDK 3.0,I run a edge detection on DSP,now I face a time problem.
The input video format is YUV422,as follow:
Y0U0Y1V0 Y2U2Y3V3 ....
But the edge detection need Y component,so I have to extract Y0Y1Y2... from the video frame Buffer by using the for loop. The conversion is more slower then the true edge detection algorithm.
A method to solve this is using DMM/tiler, I have saw the SPRUGX8 document,but I have no idea how to configure and achieve this.
Could you help me with this, give me some ideas、documents、example and other help?Thank you in advance.
My Regards,
JQ Zhao