I want to do H.264 encoding at a sub-window of the video captured by DM365EVM.
For instance, the captured image size is 1280x720, but I only want to encode 800x600 started from pixel(100,20).
How can I do this without perform a memory copy op which is of poor efficiency?
I tried to modify DM365EVM demo "encode" as following:
1. set dynParams->inputWidth to 800 and dynParams->inputHeight to 608 before invoking Venc1_create().
2. set dim.x = dim.y = 0, dim.width = 800, dim.height = 608 and call BufferGfx_setDimensions(hCapBuf, &dim) before invoking Venc1_process(hVe1, hCapBuf, hDstBuf).
Then I used GStreamer to decode the encoded file test.h264.
gst-launch filesrc location=/test/test071816.264 !\
TIViddec2 codecName=h264dec engineName=codecServer !\
queue ! tidisplaysink2 video-standard=720p \
display-output=component -v
But I didn't get the right output video.
Did I set anything wrong?
