Project Name README

Step 1:

• If custom training model, follow instructions in 'training' folder.
• If deploying pretrained model, obtain weight file from train/weed_detection/training.

Step 2:

Inference on Darknet on Jetson Nano:

1. Navigate to the darknet folder in the terminal.
2. Type the following commands:

   make
   ./darknet detector demo cfg/obj.data cfg/yolov4-tiny-custom.cfg yolov4-tiny-custom_best.weights -c 0

   • Notes:
     • Remember to check that locations of files are accurate.
     • Can also take image and video inputs as the last argument.

Inference on Darknet on TensorRT on Jetson Nano:

1. Copy weights and configurations file to tensorrt/yolo.
2. (Important) Change the name of both files to 'yolov4-tiny-416'.
3. Type the following commands in the terminal:

   cd ${HOME}/project/tensorrt_demos/ssd
   ./install_pycuda.sh
   cd ${HOME}/project/tensorrt_demos/plugins
   make
   cd ${HOME}/project/tensorrt_demos/yolo
   python3 yolo_to_onnx.py -m yolov4-416
   python3 onnx_to_tensorrt.py -m yolov4-416
   cd ${HOME}/project/tensorrt_demos
   python3 trt_yolo.py --usb 0 -m yolov4-416

   • Notes:
     • To run inference on videos or images, will need to change trt_yolo_cv.py.
     • Change cv2.VideoCapture argument.