TDA4VM: How to display own h.264 video source

Shine

Part Number: TDA4VM

Hello Champs,

SW: PROCESSOR-SDK-LINUX-SK-TDA4VM 0806010
HW: SK-TDA4VM

Customer replaced the video source with his own h.264 video, but the video can't display.

input1:
source: /opt/edgeai-test-data/videos/video_0000_h264.h264
format: h264
width: 1280
height: 720
framerate: 30
loop: True

/opt/edgeai-gst-apps/apps_python# ./app_edgeai.py ../configs/image_classification.yaml

But the code is stuck, it can't exit with ctrl+c, the only method is to power off.

Below is the log

libtidl_onnxrt_EP loaded 0x335ef270
Final number of subgraphs created are : 1, - Offloaded Nodes - 289, Total Nodes - 289
APP: Init ... !!!
MEM: Init ... !!!
MEM: Initialized DMA HEAP (fd=4) !!!
MEM: Init ... Done !!!
IPC: Init ... !!!
IPC: Init ... Done !!!
REMOTE_SERVICE: Init ... !!!
REMOTE_SERVICE: Init ... Done !!!
101.597246 s: GTC Frequency = 200 MHz
APP: Init ... Done !!!
101.600542 s: VX_ZONE_INIT:Enabled
101.600550 s: VX_ZONE_ERROR:Enabled
101.600553 s: VX_ZONE_WARNING:Enabled
101.601435 s: VX_ZONE_INIT:[tivxInitLocal:130] Initialization Done !!!
101.602555 s: VX_ZONE_INIT:[tivxHostInitLocal:93] Initialization Done for HOST !!!
)0[1;67r[m[4l[?7h[H[J
+--------------------------------------------------------------------------+
| Object Detection[3;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+
| Input Src: /opt/edgeai-test-data/videos/video_0004.h264[6;77H|
| Model Name: ONR-OD-8200-yolox-nano-lite-mmdet-coco-416x416[7;77H|
| Model Type: detection[8;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+[H[J
+--------------------------------------------------------------------------+
| Object Detection[3;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+
| Input Src: /opt/edgeai-test-data/videos/video_0004.h264[6;77H|
| Model Name: ONR-OD-8200-yolox-nano-lite-mmdet-coco-416x416[7;77H|
| Model Type: detection[8;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+[H[J
+--------------------------------------------------------------------------+
| Object Detection[3;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+
| Input Src: /opt/edgeai-test-data/videos/video_0004.h264[6;77H|
| Model Name: ONR-OD-8200-yolox-nano-lite-mmdet-coco-416x416[7;77H|
| Model Type: detection[8;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+[H[J
+--------------------------------------------------------------------------+
| Object Detection[3;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+
| Input Src: /opt/edgeai-test-data/videos/video_0004.h264[6;77H|
| Model Name: ONR-OD-8200-yolox-nano-lite-mmdet-coco-416x416[7;77H|
| Model Type: detection[8;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+[H[J
+--------------------------------------------------------------------------+
| Object Detection[3;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+
| Input Src: /opt/edgeai-test-data/videos/video_0004.h264[6;77H|
| Model Name: ONR-OD-8200-yolox-nano-lite-mmdet-coco-416x416[7;77H|
| Model Type: detection[8;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+[H[J
+--------------------------------------------------------------------------+
| Object Detection[3;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+
| Input Src: /opt/edgeai-test-data/videos/video_0004.h264[6;77H|
| Model Name: ONR-OD-8200-yolox-nano-lite-mmdet-coco-416x416[7;77H|
| Model Type: detection[8;77H|
+--------------------------------------------------------------------------+
+--------------------------------------------------------------------------+[H[J
+--------------------------------------------------------------------------+

Thanks
Regards
Shine

over 1 year ago

0 Fabiana Jaimes over 1 year ago

TI__Mastermind 19680 points

Hi Shine,

Based on the logs, it seems like the input is video_0004.h264 instead of video_0000_h264.h264. Please ensure that the correct input is selected in the flows section of the image_classification.yaml file. I recommend referencing this page to help with configuring the application.

I hope this helps.

Thank you,

Fabiana

0 Shine over 1 year ago in reply to Fabiana Jaimes

TI__Genius 13655 points

Hello Fabiana,

Thank you very much.

Fabiana Jaimes said:
it seems like the input is video_0004.h264 instead of video_0000_h264.h264.

Customer has modified the yalm file with video_0004.h264, the issue still exists.

His own video can run on other SDK version, but it can't run on the PROCESSOR-SDK-LINUX-SK-TDA4VM 08060103, is it a bug?

+1 Fabiana Jaimes over 1 year ago in reply to Shine

TI__Mastermind 19680 points

Hi Shine,

Does the width and height of the video match what is defined in the input section? Could you share the yaml file being used?

Thank you,

Fabiana

Processors

Processors forum

TDA4VM: How to display own h.264 video source