Programming Computer Vision with Python

Livro digital

Título:
Programming Computer Vision with Python

Autor:
Jan Erik Solem

Categoria:
Tecnologia > IA

Doador:
Raffaello D. N.

Sinopse:
Computer vision becomes approachable when each idea produces something visible. Jan Erik Solem begins with image handling in PIL, NumPy, SciPy, and Matplotlib, then advances through Harris corners, SIFT descriptors, homographies, panoramas, camera calibration, augmented reality, and multi-view reconstruction. Later chapters connect clustering and visual words to image search before moving into classification, segmentation, and OpenCV. The emphasis stays close to implementation. Readers learn how images become arrays, how features can be detected and matched, how camera geometry recovers structure, and how retrieval systems index and rank visual content. K-nearest neighbors, Bayesian classifiers, support vector machines, optical character recognition, graph cuts, video processing, and tracking appear as parts of a single practical toolkit rather than isolated techniques. For Python programmers who want a working introduction to the field, the draft combines foundational algorithms with reproducible examples and enough theory to explain why they work. Its greatest strength is the route from manipulating a first image to building systems that recognize, search, reconstruct, and track. The result is a hands-on foundation for further study in computer vision and machine learning.