- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi guys,
I am a bit of a newbie to programming in general and am doing a project for school. I am using Quartus and I have a DE2 board and a terasic camera which I am using to capture a video stream and do image processing on it (simple colour segmentation). Now I need to feed these filtered pixels into Nios for further manipulations so I have to reduce the resolution of the image otherwise Nios has memory issues. What I need to do is only sample every 20th pixel. I have the x and y coordinates of the pixels. So I thought I would do this in schematic capture and have a modulus counter counting up til 20 for x and y, and only taking that pixel... how would I do this? any clue? My resolution is 640x480 so my aim is to reduce this to 32x24 because that would be a much smaller array to feed into Nios. Any help would be highly appreciated because this is very urgent. Thankssss DaniLink Copied
2 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
--- Quote Start --- Hi guys, Now I need to feed these filtered pixels into Nios for further manipulations so I have to reduce the resolution of the image otherwise Nios has memory issues. What I need to do is only sample every 20th pixel. I have the x and y coordinates of the pixels. So I thought I would do this in schematic capture and have a modulus counter counting up til 20 for x and y, and only taking that pixel... how would I do this? any clue? My resolution is 640x480 so my aim is to reduce this to 32x24 because that would be a much smaller array to feed into Nios. --- Quote End --- Before actually feeding this data to the cpu, you can simply store it on an on-board memory. Later, you can sample the data from the memory either directly in NIOS or in your HDL design (from where you can feed it to NIOS).
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
[slightly off topic] The standard approach in reducing the image size - without losing information - is to low-pass filter the image first, and then downsample. You could be doing that, you just didn't mention it.

Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page