Hello, when I built a real-time target detection program on the nano, and found that when the test shot the current picture, the results would not be displayed on the external screen of the nano for a second.
Please explain more for your design.
I deployed the yolov4-tiny algorithm model on nano, and then connected an external csi camera, without using tensorrt and deepstream for acceleration. There will be a delay of about one second during the test. For example, if the camera is pointed at me now, the result will be displayed in about 1s.
Could you check without any algorithm like gst-launch-1.0 nvarguscamerasrc … or nvgstcapture-1.0
This is the sentence used when importing the video at cv2.cap. Could you please post the full version
I use this input statement, there will be a delay of about 1s
Please verify this pipeline.
gst-launch-1.0 nvarguscamerasrc ! "video/x-raw(memory:NVMM)" ! nvvidconv ! xvimagesink sync=false
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.