Input multiple images on deepstream-app

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Hi, I’m testing out the object detection deepstream-app samples. I can see that the deepstream_app_config_yolo[…].txt has a MultiURI option in the [source0] stanza. Can it be used to specify multiple images as inputs?

1 Like
2

Hi,

YES. We also have example to demonstrate image decoding instead of video.

/opt/nvidia/deepstream/deepstream-4.0/sources/apps/sample_apps/deepstream-image-decode-test/

Thanks.

3

Thanks - but the same input format doesn’t seem to work for the Yolo demo:

deepstream-app -c deepstream_app_config_yoloV3.txt -i /home/tx2/Documents/YoloV3/obj/IMG_0693.jpg

will work. But this won’t:

deepstream-app -c deepstream_app_config_yoloV3.txt -i /home/tx2/Documents/YoloV3/obj/IMG_0693.jpg /home/tx2/Documents/YoloV3/obj/IMG_0694.jpg

The config file is set to MultiURI.

4

Hi,

Sorry for the late update.
You will need to use the deepstream-image-decode-app rather than deepstream-app.

For example:

$ ./deepstream-image-decode-app <file1> [file2] ... [fileN]
$ ./deepstream-image-decode-app file1.mjpeg file2.mjpeg

Thanks.

5

How do I save the inferenced image from the deepstream-image-decode-app?

I ran the image-decode-app and it display the inferenced image for less than 1 second, and disappears ans also tried adding

output-file=out.jpg

in the config file, and nothing happend.

Can you help on this?

Thanks

1 Like
6

Hello jayrodge, have you solved this problem?

1 Like
7

Not yet @1456873666

1 Like
8

Hey man!

Have you managed to solve this ? I am also wondering about that.

Thanks.