# Tools

**Repository Path**: GhostDancer/tools

## Basic Information

- **Project Name**: Tools
- **Description**: No description available
- **Primary Language**: Python
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 2
- **Forks**: 0
- **Created**: 2022-05-04
- **Last Updated**: 2022-11-06

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Tools for our group

### Run Script
Once GPU is empty, automatically run the code.

### Parse Results from Log
Given a log file in KITTI format, we can parse 3D mAP and plot the log.
<p align="center">
  <img src="parse-results-from-log/log/log.jpg" width="100%" height="100%">
</p>

### Statistic Center Distribution
Statistic (x,y,z) location of all 3D boxes, used to calculate POINT_CLOUD_RANGE.
<p align="center">
  <img src="statistic-center-distribution/center-distribution.png" width="100%" height="100%">
</p>

### Adjust Annotations
Convert annotations into 3 categories, put DontCare annotations at last.
<p align="center">
  <img src="adjust-annotations/adjust-annotations.jpg" width="100%" height="100%">
</p>

### Statistic Box Mean Size
Statistic box mean size, used to calculate mean_size in BOX_CODER_CONFIG.

### pcd2npy
Convert lidar file from .pcd to .npy, to avoid the incompatibility between open3d and torch.

### pcd-visualizer
Visualize the dair-v2x pcd file(.pcd).
The x line show the directory of the labels.
You can use this file like follows:
```
python pcd-visualizer/pcd-visualizer.py --pcd-file 000010.pcd --label-file 000010_cameralabel.json --show-range True
```
<p align="center">
  <img src="pcd-visualizer/example.png">
</p>

### frame2gif
Visualize online action detection results in the GIF format.
<p align="center">
  <img src="frame2gif/data/frame2gif.png">
</p>