Capacitacao_Visao_Computacional
收藏资源简介:
“计算机视觉培训”数据集是一个面向教育用途的结构化资源集合,源自同名的计算机视觉培训课程。该数据集的核心是组织为一系列实践“活动”,旨在系统性地教授数字图像处理、计算机视觉和机器学习应用的基础知识与技能。 数据集包含9项详细的活动,其中5项为必修,4项为选修。这些活动涵盖了计算机视觉的多个关键领域:从图像处理基础(如分辨率、色彩空间、量化)和过滤技术(空间域与频域滤波、噪声分析),到高级主题如图像分割、形态学操作、特征提取(使用ORB、GLCM、LBP等描述符),以及与Kaggle平台结合的机器学习管道实验。每项活动通常包含任务说明(PDF)、可执行的Jupyter Notebook代码、文档和支持文件。 在形式上,数据集通过四个主要字段进行描述:活动名称(`activity`)、活动描述(`description`)、对应的数据集文件夹(`dataset_folder`)和相关的Jupyter笔记本文件序列(`notebooks`)。实现这些活动主要依赖Python技术栈,包括OpenCV、scikit-image、NumPy和Matplotlib等库,并设计为可在Google Colab或本地Jupyter环境中运行。 该数据集适用于计算机视觉和图像处理的教学、自学以及作为实践代码示例的参考。其设计目的是通过动手实践帮助学习者掌握从基础图像操作到复杂特征分析和机器学习集成的完整技能链。
The "Computer Vision Training" dataset is a structured collection of educational resources derived from the eponymous computer vision training course. At its core, this dataset is structured as a series of practical "activities" designed to systematically impart foundational knowledge and practical skills in digital image processing, computer vision, and machine learning applications. The dataset includes 9 detailed activities, among which 5 are mandatory and 4 are optional. These activities cover multiple key areas of computer vision: ranging from the fundamentals of image processing (such as resolution, color space, and quantization) and filtering techniques (spatial-domain and frequency-domain filtering, noise analysis), to advanced topics including image segmentation, morphological operations, feature extraction (using descriptors like ORB, GLCM, LBP), and machine learning pipeline experiments integrated with the Kaggle platform. Each activity typically includes task instructions (in PDF format), executable Jupyter Notebook code, documentation, and supporting files. Formally, the dataset is defined by four primary fields: `activity` for the activity name, `description` for the activity description, `dataset_folder` for the corresponding dataset directory, and `notebooks` for the sequence of associated Jupyter Notebook files. These activities are primarily implemented using the Python technology stack, including libraries such as OpenCV, scikit-image, NumPy, and Matplotlib, and are designed to run on either Google Colab or local Jupyter environments. This dataset is well-suited for teaching computer vision and image processing courses, self-directed learning, and serving as a reference for practical code examples. Its core design goal is to enable learners to master the full skill set spanning from basic image manipulation to advanced feature analysis and integrated machine learning applications via hands-on practice.
数据集概述
- 名称: Capacitação em Visão Computacional (计算机视觉培训)
- 许可证: CC BY-SA 4.0
- 语言: 葡萄牙语 (pt)
- 标签: computer-vision, image-processing, education, jupyter, python, dataset
- 数据集规模: n < 1K
- 任务类别: 图像分类、图像分割、目标检测、特征提取
数据集内容
该数据集包含计算机视觉培训课程的实践和理论活动。存储库按活动组织为文件夹,每个文件夹包含:
- 活动说明 (PDF 格式)
- Jupyter Notebook (如适用)
- 标准化的 README 文档
- 支持文件 (如 requirements.txt, venv 等)
活动列表
必选活动
- 全局、局部和纹理分析 (自建小型数据集) - 单元 03 | 第 2 章
- 图像描述符的比较分析 (直方图、统计信息、ORB、GLCM、LBP)
- 实践活动 - 单元 3 | 第 1 章
- 分割、形态学操作和特征提取流程 (Python / Google Colab)
- 图像滤波与噪声分析 - 单元 2 | 第 1 章 | 任务 3
- 频率域低通/高通滤波器;高斯噪声和椒盐噪声分析
- 图像处理基础 - 单元 1 | 第 1 章 | 任务 2
- 构建自建小型数据集;分辨率、色彩空间、量化和格式
- 平滑、去噪与边缘检测 - 单元 2 | 第 1 章 | 任务 1
- 空间滤波器 (均值、高斯、中值);边缘检测器 (Sobel、Prewitt、Canny)
可选活动
- 批判性分析:数字图像基础 - 单元 1 | 第 1 章 | 任务 1
- 深入理论分析,应用于真实计算机视觉场景
- 频率域图像分析 - 单元 2 | 第 1 章 | 任务 2
- 傅里叶变换;低通、高通和带通滤波器分析
- 分割、形态学操作与特征提取活动 - 单元 3 | 第 1 章
- 计算机视觉流程概念分析;工业场景应用
- Kaggle 上的计算机视觉实验 - 单元 03 | 第 2 章
- 探索 Kaggle 竞赛和笔记本;连接机器学习流程
使用的技术
- Python 3.8+
- Jupyter Notebook
- OpenCV (cv2)
- scikit-image
- NumPy
- Matplotlib
- Google Colab
如何获取数据集
通过 Python
python import requests, zipfile, os repo_url = "https://github.com/arvoredossaberes/Capacitacao_Visao_Computacional/archive/refs/heads/main.zip" response = requests.get(repo_url) with open(repo.zip, wb) as f: f.write(response.content) with zipfile.ZipFile(repo.zip, r) as zip_ref: zip_ref.extractall(.) os.remove(repo.zip)
通过 GitHub (HTTPS)
bash git clone https://github.com/arvoredossaberes/Capacitacao_Visao_Computacional.git
或: bash wget https://github.com/arvoredossaberes/Capacitacao_Visao_Computacional/archive/refs/heads/main.zip unzip main.zip
通过 GitHub (SSH)
bash git clone git@github.com:arvoredossaberes/Capacitacao_Visao_Computacional.git
通过 Hugging Face
python from huggingface_hub import snapshot_download repo_id = "arvoredossaberes/Capacitacao_Visao_Computacional" local_dir = "./Capacitacao_Visao_Computacional" snapshot_download(repo_id=repo_id, local_dir=local_dir, repo_type="dataset", local_dir_use_symlinks=False)
或使用 CLI: bash pip install huggingface_hub huggingface-cli download arvoredossaberes/Capacitacao_Visao_Computacional --repo-type dataset --local-dir ./Capacitacao_Visao_Computacional
数据集结构 (特征)
| 特征名 | 类型 | 描述 |
|---|---|---|
| activity | string | 活动名称 |
| description | string | 活动描述 |
| dataset_folder | string | 数据集文件夹路径 |
| notebooks | sequence[string] | 关联的 Jupyter Notebook 路径列表 |
其他信息
- 创建日期: 2026-05-10
- 版本: 1.1
- 最后更新: 2026-05-10




