Hi,
I am trying to implement a 8 connected components in dm6437. I developed the algorithm with Pc and output works fine. but when port it into dm6437 it takes on avg 100 ms(depends on number of components) for a single frame. my frame widthxheight is approx 1000x600. Since this is a sequential algorithm,loops are not getting pipelined as i can see the message ,SP disqualified due to control code. i need the code to work under 20 ms as this is just part of my algorithm. is there a way to optimise it or should i do for another processor. I have been looking into c6000 optimisation workshop lately but nothing has changed much..