Motion - Feature Request 2014x 04x 10x 134646

Feature Request: Raspberry Pi GPU Optimization


Raspberry Pi is gpu powerful devices but the cpu is equievalent to pentium 2 CPU is the bottleneck

So motion software need improvement to GPU Optimization for raspberry pi bcm2835

The example is raspivid , raspberry pi camera recording which implemented GPU Optimization

we except motion software can implent GPU Optimization on raspberry pi gpu , cpu armv6 That will make raspberry pi more less cpu work for this software

-- IrvanKro - 10 Apr 2014

Follow up


I would like to see this as well

-- MortenBirkelund - 25 May 2014

I would like to work on something like that, but I never dealt with embedded GPUs, so it is a bit complicated. It should not be a huge problem as the (I guess open source?) raspivid implemented that, firstly for the Raspberry Pi camera.

-- JohnMillers - 01 Aug 2014

Bump! Seems common sense to do this.

-- MarkMagness - 07 Jan 2015

As far is I know, using the GPU on the Raspberry Pi is rather difficult because of the GPU's closed nature. Nevertheless the Raspberry delivers great performence when encoding/decoding H264. That's why the Raspberry can record 1080p videos with 30fps.

When using the Raspicam there is the opportunity to get the motion vectors of a frame instead of the full resolution pixel image, which can be compared to the vector of the next frame. With this the perfomance of the motion detection is greatly improved, as not every pixel is compared! An example can be found and

Another posibillity for general purpose USB/Network cams would be to allow motion detection on one camera but recording on another other one (using the same device with two diffeent resolutions would be even better). This setup can be used to do motion detection on a low-resolution image but record on the 1080p device.

-- JuergenHausladen - 13 Mar 2015
Topic revision: r5 - 13 Mar 2015, JuergenHausladen
Copyright © 1999-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Please do not email Kenneth for support questions (read why). Use the Support Requests page or join the Mailing List.
This website only use harmless session cookies. See Cookie Policy for details. By using this website you accept the use of these cookies.