Part Number: TDA2EXEVM
Hi,
I am using OV10635 capture sensor, and using TI deep learning object detection.
The input resolution of capture sensor are 1280x720 .
From the chains_tidlOD.c , i know the capture link will do the resize if my input of model is smaller
ChainsCommon_SingleCam_SetCapturePrms(&pUcObj->CapturePrm,
CAPTURE_SENSOR_WIDTH,
CAPTURE_SENSOR_HEIGHT,
DEC_OUT_WIDTH,
DEC_OUT_HEIGHT,
pObj->chainsCfg->captureSrc
);
but if the input width and height of my model are larger ,for example ,1024x1024 .
does capture link will do the interpolation for us?(1280x720 to 1024 x1024)
Best Regards,
Eric Lai
