
Computer Vision: Algorithms and Applications by Richard Szeliski is an authoritative manual on image processing, 3D reconstruction, and visual deep learning. The field combines mathematical image processing with modern convolutional networks and Vision Transformers. Szeliski provides an encyclopedic exploration of computational geometry, feature recognition, image matching, and neural rendering.
What you will learn
- Formulate image processing pipelines using spatial filtering, Fourier transforms, and wavelets
- Extract invariant local image features using SIFT, ORB, and learned neural representations
- Reconstruct 3D scene geometry using structure from motion and multi-view stereo
- Compute dense optical flow and track objects across complex video sequences
- Implement semantic segmentation, object detection, and instance recognition with CNNs
- Explore novel view synthesis, neural radiance fields (NeRF), and generative image models
Who this is for
Computer vision researchers, machine learning engineers, and graduate students building visual perception systems. Strong mathematical foundations in linear algebra, multivariable calculus, and probability are assumed. Prior computer vision experience is not required.


