Cabbage Project

---

Overview

This project detects cabbages in video frames and generates maps and annotated outputs.
It supports both TensorRT FP16 inference and PyTorch saved models.

---

Setup and Usage

Activate your environment and run everything with these commands:

# 1. Create and activate the Conda environment
cd recreate_env
chmod +x recreate_env.sh
./recreate_env.sh

# 2. Create the TensorRT FP16 engine
bash scripts/create_TensorRT.sh

# 3. Run the TensorRT inference script to generate outputs
python main.py

Output

Running the scripts will generate the following files:

• .mp4 — Video with cabbage annotations

• .html — Map of detected cabbages

• .txt — Timing and related information