Not able to see video in demo

[gangooteli]

In People-counter demo, I am not able to see visual. In object detection, there is not option to pass input as a camera. Only path of video can be passed.

Below is the command I am running. ./obj_recognition -i Pedestrain_Detect_2_1_1.mp4 -m /opt/intel/computer_vision_sdk/deployment_tools/intel_models/person-detection-retail-0013/FP32/person-detection-retail-0013.xml -l /opt/intel/computer_vision_sdk/deployment_tools/intel_models/person-detection-retail-0013/FP32/person-detection-retail-0013.bin -t SSD -thresh 0.7 -d CPU 0 2>/dev/null | ffmpeg -v warning -f rawvideo -pixel_format bgr24 -video_size 544x320 -i - http://localhost:8090/fac.ffm

I am only able to see people count. The processed video is not visible.

Running without argument i gives below error: ERROR: image(s) for inference - not set

Can we run this demo for camera setup? Please help !!

[Propanu]

Hi @gangooteli, if I remember correctly you can pass a v4l camera device number to the input flag, e.g. -i 0 with this demo. The video not displaying seems like an issue with ffserver. Try opening the stream directly with ffplay or vlc to see if this is indeed a server configuration issue or something to do with the browser.

[saket424]

For v4l2 the following syntax worked ./obj_recognition -i /dev/video0 -m /opt/intel/computer_vision_sdk/deployment_tools/intel_models/person-detection-retail-0013/FP32/person-detection-retail-0013.xml -l /opt/intel/computer_vision_sdk/deployment_tools/intel_models/person-detection-retail-0013/FP32/person-detection-retail-0013.bin -t SSD -thresh 0.7 -d CPU 0 2>/dev/null | ffmpeg -v warning -f rawvideo -pixel_format bgr24 -video_size 544x320 -i - http://localhost:8090/fac.ffm