tensorrt_car_det_yolov3

1. Install TensorRT on Ubuntu

2. Test TensorRT_yolo3_module

• a. Download vehicle / license plate detection model from Baidu online disk
• b. python2 yolov3_to_onnx.py. Convert yolov3 model .weights file to .onnx file
• c1. python3 onnx_to_trt_1batch.py. If you only need to process one image each time, for example you only have one camera. Executing this script you need python 3.x, and you will have a file named yolov3-608.trt, which is the file we ultimately need.
• c2. python3 onnx_to_trt_multibatch.py. If you need to process multiple images each time, for example you have multiple cameras. Executing this script you also need python 3.x, and you will have a file named yolov3-608.trt, which is the file we ultimately need. And the data accuracy is FP16, so the acceleration is more obvious.
• d1.python3 trt_yolo3_module_1batch.py, if you choose c1
• d2.python3 trt_yolo3_module_multibatch.py,if you choose c2. It detects 4 images at a time.

3. Test yolov3-tiny-onnx-TensorRT

• a. python2 yolov3_tiny_to_onnx.py