Altera_Forum
Honored Contributor
15 years agoComputer vision
I'm working on a machine vision project where I have a high data rate (>100 MB/s) to process. The camera will be connected to the system using either GigE vision or CameraLink protocol, and let's assume for now that I have to do a simple Sobel edge detection on the images. I have the image processing code already in C and Matlab. I want to process these pixels in real time and I was told to try this on FPGAs
I have zero experience using FPGAs and so I would like some tips from you on how to get started with this task. Which tools do I have to use? I have access to a full version of Quartus and Matlab with Simulink. I have a DE2 education program board but we can buy a higher spec one if needed. I've seen NiosII and SOPC, and also C-to-Hardware Acceleration Compiler being mentionned several times. Is that the way to proceed for a complete newbie in HDLs, and if yes can you send me links to tutorials or working examples? Are these tools already included in Quartus or do I need a separate license? What are the limitations of using this direct C language translation to HDL languages? And what is the preferred HDL to use? I've also seen Mathworks' Simulink HDL coder and DSP builder also coming up in a few forums, and all these tools are just confusing me evern more. I would be very grateful for any advice on which direction to take. Thanks PS: Please accept my apologies if similar newbie questions have been posted before, but I couldn't get a clear answer to what I wanted by reading other posts.