I am new to Xilinx. I started looking at xfOpenCV functions while I noticed that performance estimation of functions looked strange for me.
For example, if we look at WarpAffine function at https://www.xilinx.com/html_docs/xilinx2018_3/sdsoc_doc/xfopencv-library-api-reference-ycb1504034263746.html#dbh1504034283793 we see that Latency estimate is 19.3 ms (max) for 1 pixel computation. Does this mean it takes maximum 19.3 ms to complete function? It looks strange for me because function like WarpAffine should not take too much time to complete.
Please, correct me if I am taking this wrong.