Hi, I am doing a project on pedestrian detection using image
processing from a liive video feed.
I would really like to know what the best product for this will be? Will it be a DSP development board or FPGA? And also if it is possible what part number.
The costs will not matter that much, but it should not be way to expensive.
Any advise is helpful, thank you in advanced.
Hi Savva,
Savva Antoniadis wrote:
Hi, I am doing a project on pedestrian detection using image
I would really like to know what the best product for this will be? Will it be a DSP development board or FPGA? And also if it is possible what part number.
The costs will not matter that much, but it should not be way to expensive.
Any advise is helpful, thank you in advanced.
It really depends on your final application requirments, ie power requirements, time requirements, amount of effort wanted to develop the product.
It would be best to start using openCV on a full PC to figure out which algorithms work best for your applcation. You can quickly tune your algorithm to fit your needs without all of the development overhead of creating custom hardware in an FPGA.
Once you have figured out your required algorithm, processing intensity of application you can decide which platform you would like to use for the end applcation. If not very processing intensive you may be able to get away with a Rpi or BBB running openCV. But, you may need to port some of the heavy lifting to custom hardware on an FPGA. You can see Jonathan Piat's blog posts on running machine vision on an FPGA to get feel for what it might take to use an FPGA for your application.
Gradient Filter implementation on an FPGA - Part 1 Interfacing an FPGA with a camera
Gradient Filter implementation on FPGA : Part 2 Implementing gradient Filter
Anyhow good luck and keep us posted on your progress and design decision making!
