This course covers the fundamental principles and methodologies of computer vision, with emphasis on modern AI-based visual perception systems. Topics include image formation and photometric modeling; camera geometry and calibration; light and shading; image filtering and resampling; image pyramids; edge detection; feature detection and matching; image alignment and robust estimation (RANSAC); depth recovery from stereo; multi-view geometry; structure from motion; motion estimation and optical flow. The course further covers learning-based image understanding, including convolutional neural networks, vision transformers, visual representation learning (self-supervised / contrastive learning), image classification, object detection, and image segmentation. Advanced topics include open-vocabulary recognition, vision-language models, depth estimation, 3D reconstruction, and deep learning methods for 3D shape and scene analysis using multi-view, point-based, volumetric, and graph-based representations.