Wishlist
Book Summary
Multiple View Geometry in Computer Vision, Second Edition, by Richard Hartley and Andrew Zisserman is the definitive technical reference on geometric methods for reconstructing 3D scenes from multiple camera images. It develops projective geometry, camera models, epipolar constraints, and bundle adjustment with rigorous mathematics and practical algorithms. Graduate students and researchers in computer vision, robotics, and photogrammetry rely on it as both a textbook and lasting reference.
This book is not available
Store  |
Condition |
Deals & Coupons | Price |
Est. Shipping |
Total |
Buy |
|---|---|---|---|---|---|---|
| Sorry we could not find any result for “9780521540513” | ||||||
About Multiple View Geom Comp Vision 2ed
This book systematically develops the geometry underlying multi-view computer vision, beginning with projective geometry, transformations, and camera calibration before advancing to epipolar geometry, the fundamental and essential matrices, and triangulation. Later chapters cover structure from motion, bundle adjustment, auto-calibration, dense matching, and reconstruction from video sequences. Hartley and Zisserman combine formal derivations with algorithmic recipes, numerical considerations, and implementation guidance, so readers can move from theory to working code. The second edition expands treatment of important developments and adds substantial new material. Its enduring value lies in unifying classical projective geometry with modern estimation techniques, giving readers a coherent framework for understanding how 3D structure and camera motion are recovered from image correspondences.
Who should read Multiple View Geom Comp Vision 2ed?
Graduate students and researchers in computer vision, robotics, photogrammetry, and augmented reality will benefit most, especially those implementing structure-from-motion or visual SLAM systems. Practitioners with strong linear algebra and probability backgrounds gain a rigorous foundation for multi-view reconstruction. It also suits engineers seeking a durable reference rather than a tutorial, and instructors building advanced courses. Readers without mathematical maturity may find the density challenging, so prior exposure to geometry and estimation is helpful.
Pros & Cons
Pros
- ✓ Authoritative, comprehensive treatment of multi-view geometry fundamentals
- ✓ Rigorous mathematical derivations paired with practical algorithms
- ✓ Extensive coverage from camera models through bundle adjustment
- ✓ Clear notation and consistent framework across chapters
- ✓ Widely cited reference valuable for research and implementation
Cons
- ✗ Dense mathematical style demands strong linear algebra background
- ✗ Some material feels dated relative to deep-learning approaches
- ✗ Limited hands-on code and modern software examples
- ✗ Assumes prior familiarity with estimation and optimization
- ✗ Not an introductory text for beginners in computer vision
Similar books
Frequently asked questions
Is this book suitable for self-study?
Yes, motivated readers with solid linear algebra and probability can study it independently, though the exercises and density reward persistence. Pairing it with a course or implementation project makes the material much more accessible.
Does the second edition differ substantially from the first?
The second edition adds significant new material and reorganizes several topics, making it the preferred version for most readers. It remains the standard citation in the field.
What mathematical background is required?
Readers should be comfortable with linear algebra, matrix decompositions, probability, and basic optimization. Some familiarity with projective geometry helps but is developed within the text.
Is the book still relevant given deep learning advances?
Yes, its geometric foundations remain essential for calibration, structure from motion, and SLAM. Modern learning-based systems still rely on these classical formulations.
Does it include algorithms and implementation guidance?
Yes, chapters present algorithmic summaries and discuss numerical and implementation issues. It is not a code repository, but it supports practical implementation well.
At Windsorbooks, we compare prices from trusted sellers so you can buy, rent, or sell textbooks at the best total price — shipping included.
As an Amazon Associate, Windsorbooks.com may earn a small commission on qualifying purchases.