Brazilian_Road_Signs_Dataset

# Brazilian Road Signs Dataset
*This dataset contains high-quality images of Brazilian road and traffic signs collected from various urban and rural environments. It supports AI research in computer vision, object detection, and autonomous driving systems adapted to Brazil’s signage standards and language.*

## Contact
For queries or collaborations related to this dataset, contact:
- [e-mail]
- [e-mail]

## Supported Tasks

- **Task Categories**:
- Image Classification
- Object Detection
- Scene Understanding

- **Supported Tasks**:
- Recognition and classification of Brazilian traffic signs
- Detection of multilingual and symbolic road indicators
- Scene segmentation for navigation and self-driving datasets
- Visual understanding for smart transportation and ADAS (Advanced Driver Assistance Systems)
- Safety compliance and infrastructure mapping

## Languages

- **Primary Language**: Portuguese (used in textual road signs)
- **Secondary Presence**: Universal symbols (non-linguistic signs)

## Dataset Creation

### Curation Rationale
This dataset was created to advance AI capabilities in road safety, automated navigation, and urban planning. It supports visual recognition of traffic elements specific to Brazil’s highway code and driving environments.

### Source Data
- **Contributors**: Field data collection and publicly available road imagery under open licenses
- **Collection Process**: Images were captured from highways, urban streets, and rural areas across Brazil. Each sign was categorized by type (regulatory, warning, informational, or guide). Personally identifiable data (faces, license plates) was blurred or removed.

### Other Known Limitations
- **Bias**: Overrepresentation of urban traffic signs compared to rural or regional variants
- **Weather and Lighting Variability**: Conditions may affect clarity and detection accuracy
- **Geographical Coverage**: Certain states or remote roads may have limited representation

## Intended Uses

### ✅ Direct Use
- Training computer vision models for traffic sign recognition
- Research in autonomous driving and ADAS development
- Road infrastructure mapping and visual classification
- Traffic monitoring and safety compliance analysis
- Academic work on multilingual and symbolic sign understanding

### ❌ Out-of-Scope Use
- Real-time surveillance or tracking of individuals or vehicles
- Misuse for commercial mapping without authorization
- Use in law enforcement decision-making without human review

## License

CC BY 4.0

# 巴西道路标志数据集
*本数据集收录了从各类城乡环境中采集的高质量巴西道路与交通标志图像，可支撑适配巴西标志规范与语言的计算机视觉、目标检测及自动驾驶系统相关AI研究。*

## 联系方式
若您有关于本数据集的咨询或合作意向，请联系：
- [e-mail]
- [e-mail]

## 支持任务
- **任务类别**：
- 图像分类
- 目标检测
- 场景理解

- **支持的具体任务**：
- 巴西交通标志的识别与分类
- 多语言与符号化道路标识的检测
- 面向导航与自动驾驶数据集的场景分割
- 智能交通与ADAS（高级驾驶辅助系统，Advanced Driver Assistance Systems）的视觉理解
- 安全合规性与基础设施测绘

## 语言类型
- **主要语言**：葡萄牙语（应用于文字类道路标志）
- **次要存在形式**：通用符号（非语言类标志）

## 数据集构建
### 构建依据
本数据集旨在提升道路安全、自动导航与城市规划领域的AI能力，可支撑针对巴西道路交通法规与驾驶场景特有的交通元素的视觉识别研究。

### 源数据
- **贡献方**：实地采集数据及开放许可下的公开道路图像
- **采集流程**：图像采集自巴西境内的高速公路、城市街道与乡村路段。所有标志均按类型（监管类、警告类、信息类或导向类）进行分类。个人可识别信息（如人脸、车牌）已做模糊处理或移除。

### 已知局限性
- **偏差问题**：与乡村或区域变体标志相比，城市交通标志的占比偏高
- **天气与光照变化**：环境条件可能影响图像清晰度与检测精度
- **地理覆盖范围**：部分州份或偏远道路的样本量可能有限

## 预期用途
### ✅ 直接用途
- 训练用于交通标志识别的计算机视觉模型
- 自动驾驶与ADAS（高级驾驶辅助系统）开发相关研究
- 道路基础设施测绘与视觉分类
- 交通监测与安全合规性分析
- 多语言与符号化标志理解相关学术研究

### ❌ 超出适用范围的用途
- 对个人或车辆进行实时监控或追踪
- 未经授权用于商业测绘
- 未经人工审核即用于执法决策

## 授权协议
CC BY 4.0