3D数据科学与Python（影印版）
3D数据科学与Python（影印版）
Florent Poux 著
出版时间：2025年07月
页数：658
“我很少见到一本书能把复杂的3D数据工作流程讲得如此易懂。对于真正想掌握3D技术的人来说，这是一本必备之作。”
——Tarun Narayanan Venkatachalam
AI工程师，Unstructured
“这本书可以说为数据科学添加了一个全新的维度。如果你可以用3D思考，又何必满足于平面的认知呢？”
——Lipi Deepaakshi Patnaik
高级软件开发工程师，Zeta

我们的物理世界是建立在三维空间之中的。为了创造能够理解并与之交互的技术，我们的数据也必须是三维的。这本实用指南为数据科学家、工程师、研究人员提供了使用Python处理3D数据的实践方法。从3D重建到3D深度学习技术，你将学习如何从海量数据集中提取有价值的洞察，包括点云、体素、3D CAD模型、网格、图像等。
Florent Poux博士将帮助你借助前沿算法和空间AI模型的潜力，开发以自动化为核心的生产就绪系统（production-ready system）。通过本书，你将习得3D数据科学的知识与代码，实现以下目标：
● 理解3D数据的核心概念和表示方法
● 使用强大的Python库加载、操作、分析和可视化3D数据
● 应用先进的AI算法进行3D模式识别（包括监督与非监督方法）
● 使用3D重建技术生成3D数据集
● 实现自动化3D建模与生成式AI工作流
● 探索在计算机视觉/图形、地理空间情报、科学计算、机器人技术、自动驾驶等领域的实际应用
● 构建服务于空间AI解决方案的精准数字环境
  1. Foreword
  2. Preface
  3. 1. Introduction to 3D Data Science
  4. 3D Data Science in Brief
  5. 3D Data: Fundamental Building Blocks
  6. The 3D Data Science Modular Workflow
  7. 3D Data Science in the Industry
  8. Summary
  9. 2. Resources and Software Essentials
  10. Fundamental Resources
  11. Hardware Recommendations for 3D
  12. Essential Software and Tools for 3D
  13. Summary
  14. 3. 3D Python and 3D Data Setup
  15. 3D Python Setup and Libraries
  16. Creating a 3D Python Program
  17. 3D Reconstruction Methods
  18. 3D Dataset: Curation
  19. Summary
  20. 4. 3D Data Representation and Structuration
  21. 3D Data Representations
  22. 3D Data Canonical Link
  23. 3D Data Structures: k-d Trees, Octrees, BVH
  24. Summary
  25. 5. Developing a Multimodal 3D Viewer with Python
  26. 3D Python and Code Setup
  27. 3D Data Curation
  28. 3D Data Preparation
  29. Multimodal 3D Experience
  30. Summary
  31. 6. Point Cloud Data Engineering
  32. Fundamentals
  33. Strategies for Point Cloud Feature Extraction
  34. Principal Component Analysis
  35. 3D Data Registration: Unifying Perspectives
  36. Summary
  37. 7. Building 3D Analytical Apps
  38. 3D Project Environment Preparation
  39. 3D Data Fundamentals with PyVista
  40. 3D Data Structure Creation (KDTree)
  41. Covariance Matrix, Eigenvalues, and Eigenvectors
  42. Planarity, Linearity, Omnivariance, Verticality, Normals
  43. Neighborhood Definition and Selection
  44. Automation and Scaling
  45. Interactive Thresholding
  46. 3D Data Results Export
  47. Summary
  48. 8. 3D Data Analysis
  49. Types of 3D Data Analysis
  50. 3D Data Analytical Tools
  51. 3D Diagnostic Tools
  52. Summary
  53. 9. 3D Shape Recognition
  54. RANSAC from Scratch: 3D Planar Shape Recognition
  55. Region Growing for 3D Shape Detection
  56. A Hybrid Approach: RANSAC and Region Growing
  57. Summary
  58. 10. 3D Modeling: Advanced Techniques
  59. High-Fidelity Meshing
  60. 3D Voxels and Voxelization
  61. Parametric Modeling
  62. Monocular Image-based 3D Modeling: Depth Estimation and Reconstruction
  63. Summary
  64. 11. 3D Building Reconstruction from LiDAR Data
  65. Phase 1: 3D Python Setup
  66. Phase 2: Data Preparation
  67. Phase 3: Experiments
  68. Phase 4: Automation and Scaling
  69. Summary
  70. 12. 3D Machine Learning: Clustering
  71. Clustering for Unsupervised Segmentation
  72. k-Means Clustering
  73. DBSCAN for Unsupervised Segmentation
  74. Summary
  75. 13. Graphs and Foundation Models for Unsupervised Segmentation
  76. Connectivity-based Clustering
  77. The Segment Anything Model
  78. Summary
  79. 14. Supervised 3D Machine Learning Fundamentals
  80. From Unsupervised to Supervised Learning
  81. 3D Point Cloud Semantic Segmentation
  82. Specializing 3D Machine Learning with 3D Deep Learning
  83. Summary
  84. 15. 3D Deep Learning with PyTorch
  85. 3D Deep Learning Backbone
  86. Implementation with PyTorch
  87. 3D Deep Learning: The Architectures
  88. 3D Machine Learning Versus 3D Deep Learning
  89. Fine-Tuning, Transfer Learning, and 3D Data Augmentation
  90. Summary
  91. 16. PointNet for 3D Object Classification
  92. PointNet: A Point-based 3D Deep Learning Architecture
  93. 3D Object Classification
  94. Large-Scale Semantic Segmentation Considerations
  95. Summary
  96. 17. The 3D Data Science Workflow
  97. 3D Data Acquisition
  98. 3D Data Preparation and Engineering
  99. 3D Data Modeling
  100. Semantic Extraction
  101. 3D Data Visualization and Analysis
  102. Summary
  103. 18. From 3D Generative AI to Spatial AI
  104. Advanced 3D Projects
  105. Spatial AI: The Future of 3D Experiences
  106. Conclusion
  107. Index
书名：3D数据科学与Python（影印版）
作者：Florent Poux 著
国内出版社：东南大学出版社
出版时间：2025年07月
页数：658
书号：978-7-5766-2003-0
原版书书名：3D Data Science with Python
原版书出版商：O'Reilly Media
Florent Poux
 
Florent Poux是3D数据科学领域的知名专家，常年在欧洲顶尖高校从事教学与研究工作。他还是3D地理数据学院（3D Geodata Academy）的首席教授以及法国Tech 120企业的创新总监。
 
 
The animal on the cover of 3D Data Science with Python is a blue viper (Trimeresurus insularis). Also known as the white-lipped island pit viper, blue vipers are closely related to the white-lipped pit viper.
As the name suggests, blue vipers are known for their vibrant blue scales. This makes them extremely rare, as most of the snakes in this species are green. The scales can range from a bright, electric blue to a pale, powdery blue and often have an iridescent or metallic appearance.
Blue vipers are native to Komodo Island but can be found in other areas of southeast Asia as well. They prefer forested areas such as monsoon forests, bushlands, and bamboo forests where their blue coloration helps them blend in with their surroundings. Blue vipers are arboreal (tree dwellers) and nocturnal, waiting until nighttime to hunt their prey (rodents, birds, lizards, frogs, and other small mammals).
Although not generally aggressive, blue vipers will fight when provoked. They have hollow fangs which they use to inject venom into their prey. These snakes are poisonous to humans, but there are antidotes.
购买选项
定价：188.00元
书号：978-7-5766-2003-0
出版社：东南大学出版社
联系出版社邮购

© 2026，奥莱利技术咨询(北京)有限公司。版权所有。
传真: +86-10-88097463
 京ICP备05003502号