Deepstack picture resolution

[Cameraguy]

Is there a way to adjust the resolution of the images that are sent to deepstack?

 

[wittaj]

Up the bitrate and resolution of the substream or use mainstream only IF you have a powerful enough machine.

 

[Cameraguy]

Up the bitrate and resolution of the substream or use mainstream only IF you have a powerful enough machine.

So it's definitely pulling image from sub stream if they are enabled

 

[wittaj]

Yes, which is problematic for some field of view for faces. For almost anything else it is ok within reason.

 

[Cameraguy]

Yes, which is problematic for some field of view for faces. For almost anything else it is ok within reason.

I might have to pick and choose which cameras use sub stream from now on. One that detect alot of people and faces, maybe I'll go to main stream

 

[nickh66]

Perhaps this will help.
Buried in the "camera settings>trigger tab>motion sensor, configure & advanced, there is a 'high definition' option.
Quote from the help pdf. "By default, to save CPU and smooth-out noise, the image is reduced by considering it in
blocks. The High definition option actually increases the number of motion detection blocks
that are used by typically 4x."

 

[Spuds]

I use Trigger, when Triggered, High Res JPEG files ... that is what gets sent to DS to analyze. You should not have to mess with your sub or main streams,

 

[JNDATHP]

I use Trigger, when Triggered, High Res JPEG files ... that is what gets sent to DS to analyze. You should not have to mess with your sub or main streams,

Would you please post a screen shot? Thanks.

 

[Spuds]

There is what I use .... When an alert image is captured, you then may run Immediate actions prior to confirming the alert with DS For me its a 3840x2176 image, example attached.

 

[cb8]

On thing to keep in mind is that DeepStack will scale the image to the required input size of the model, that being 256x256, 416x416 or 640x640 depending on whether you have selected low, medium or high settings in DeepStack. Using the sub-stream as input to DeepStack therefore makes sense as it already matches the required input size fairly closely, while using the main stream wastes significant resource to first decode the main stream and the downsize the image afterwards. After an object has been identified though, saving hi-res alerts is certainly appealing.