How should I use the USB camera to detect in real time with this repo???

ceccocats / tkDNN

Deep neural network library and toolkit to do high performace inference on NVIDIA jetson platforms

GNU General Public License v2.0

718 stars 209 forks source link

How should I use the USB camera to detect in real time with this repo??? #219

Closed qrzeffort closed 3 years ago

qrzeffort commented 3 years ago

According to the issue#96 I have tried to use dev/video0 to substitute yolo_test.mp4, but it did not run successfully. And how to modify demo.cpp to use the USB camera like the original YoloV4 code? Like this: darknet.exe detector demo cfg/coco.data cfg/yolov4.cfg yolov4.weights -c 0 -->./demo yolo4_fp32.rt -c 0 y

ps:"'path-to-video' is the path to a video file or a camera input",Could you give an example of the "camera input"? It is not explained in the readme.md.Thank you~

SokPhanith commented 3 years ago

if you want detect video : ./demo yolo4_fp32.rt test.mp4 y if you want detect camera (jetson board) : ./demo yolo4_fp32.rt /dev/video0 y

but something you need change port camera like /dev/video1 , /dev/video2.
check camera on jetson board : ls /dev/video* if you want detect camera (windows) ./demo yolo4_fp32.rt 0 y port camera (0,1,2)