I have questions with regards to the Jetson tk1
-
From this link http://docs.nvidia.com/gameworks/index.html#technologies/mobile/opencv_known_issues.htm%3FTocPath%3DTechnologies
It seems that OpenCV4Tegra has all the functions included as if they are available from the regular OpenCV, or is it only some of the functions optimized ( for NEON SIMD and multi threading as well for for the CUDA) and other functions are available as it ( from the normal OpenCV) OR it has optimized all functions of the OpenCV? -
It looks like the OpenCV4Tergra is closed source. Is there any way to obtain the source code so that depending on the need, OpenCV4Tegra functions can be re-factored to achieve further performance enhancement using the CUDA parallel programming, NVIDIA performance primitives, vision works ( or the optimization has been already done with such tools/libraries)
-
For Jetson tk1 there are two OpenCv packages available namely the OpenCV4Tegra and another OpenCV4Tegra-python. I haven’t programmed OpenCV with Python, but in general sense, it offers the python warppers to the OpenCV functions coded in the C/C++. I’m wondering if I were to use the openCV-python package, will I get the similar results as compared with OpenCV4Tegra?
Please post your response.
Thanks in advance!