Sensor Fusion for Firefighters. Displaying heads-up video on the live feed

RE: Sensor Fusion for Firefighters. Displaying heads-up video on the live feed

dang74 — Tue, 06 Dec 2022 01:34:01 GMT

I imagine the embedded logic analyzer IP blocks are a very powerful tool for debugging. I don't have a lot of experience with Xilinx but I use the Intel /ALTERA Signal Tap II tool all the time which provides an embedded logic analyzer function.

By the way the video you posted below in the messages is very cool. Reminds me of Terminator 2.

RE: Sensor Fusion for Firefighters. Displaying heads-up video on the live feed

javagoza — Sat, 26 Nov 2022 22:20:02 GMT

It is already going from 70% LUT usage to 61% by optimizing the HLS code of the Head-up Display IP module.

RE: Sensor Fusion for Firefighters. Displaying heads-up video on the live feed

javagoza — Wed, 23 Nov 2022 21:47:50 GMT

[View:https://youtu.be/ZNtoNa4JA9k:640:360]

HDMI Heads Up Display (HUD) with 16 seven-segment virtual display modules on the AMD Xilinx SP701 development board.

In the end removing all the "optimizations" that I had introduced and going back to a cruder design I already have the 16 seven segment display modules working superimposed on the actual video feed.

RE: Sensor Fusion for Firefighters. Displaying heads-up video on the live feed

saadtiwana_int — Tue, 22 Nov 2022 08:27:41 GMT

Btw, I noticed you're already at 70%+ utilization on your LUTs. I am wondering which IPs are your major contributors to this high resource consumption? My guess would be that your VPSS is taking up a huge chunk. Unless you REALLY need it, you can remove it, or use the minimum configuration for it that you need. It's an HLS based IP, which tend to consume more resources than their pure-HDL counterparts.

I have experienced that once you start touching 80% utilization, place and route can start to fail as there are no more slices available.

RE: Sensor Fusion for Firefighters. Displaying heads-up video on the live feed

saadtiwana_int — Mon, 21 Nov 2022 12:29:48 GMT

You're getting there fast! [emoticon:c4563cd7d5574777a71c318021cbbcc8]