Niet blij met je aankoop? Geeft niet! Je kunt artikelen tot 30 dagen retourneren
Met een cadeaubon zit je altijd goed. De ontvanger kan de cadeaubon voor alles uit ons assortiment inwisselen.
Tot 30 dagen retourrecht
Few technologies have changed as quickly, or as visibly, as computer vision. A little over a decade ago, recognising a thousand everyday object categories in photographs was a research frontier; today the same capability runs on a mobile phone, and systems that segment arbitrary objects from a short text prompt, read medical scans, guide vehicles and inspect factory products are routine engineering components. Yet the speed of progress has left a gap. Researchers publish new architectures every week, practitioners assemble pipelines from pre-trained models, and students are often asked to learn either the classical theory of image formation and geometry or the latest deep networks, but rarely both in one coherent account. This book was written to close that gap.
Computer Vision Engineering treats the field as an engineering discipline. An engineer must understand how light becomes pixels, how those pixels can be cleaned and analysed with well-founded mathematics, how learned models extract meaning from them, and how the resulting systems are specified, evaluated, deployed and maintained in the real world. The book therefore follows the complete life cycle of a vision system: from image acquisition and digital image processing, through feature extraction and geometry, to deep learning for visual recognition, object detection and image segmentation, and finally to data engineering, optimisation for edge devices, monitoring, ethics and robustness.
Who this book is for
The text is intended for senior undergraduate and graduate students of computer science, electronics and communication, electrical, mechanical, biomedical and data-science programmes, and for practising engineers who wish to move from using libraries to understanding them. A working knowledge of linear algebra, calculus, probability and basic programming in Python is assumed; Appendix A reviews the mathematics that is used most often. No prior exposure to image processing or neural networks is required.
Approach and philosophy
Four principles guided the writing.
Hoi! Ik ben Libroamiko, je boekadviseur.
Hoe kan ik je helpen?