Part I Introduction 1 Overview and Contributions 2 Developments in Mobile Robot Localization Research 3 A Computer Vision System for Visual Perception in Unknown Environments Part II Unsupervised Learning 4 Theory: Clustering 5 Algorithm I: A Fast Approximate EMST Algorithm for High-Dimensional Image Data 6 Algorithm II: An Efficient K-medoids Clustering Algorithm for Large Scale Image Data 7 Algorithm III: Enhancing Complete Linkage Clustering via Boundary Point Detection 8 Algorithm IV: A New Fast k-Nearest Neighbor-Based Clustering Algorithm Part III Supervised Learning and Semi-Supervised Learning 9 Theory: K-nearest Neighbor Classifiers 10 Application I: A Fast Image Retrieval Method Based on Quantization Tree 11 Application II: A Fast Incremental Spectral Clustering Algorithm for Image Segmentation Part IV Reinforcement Learning 12 Theory: Human-Like Localization Inspired by a Hippocampal Memory Mechanism 13 Application I: A Developmental Robotic Paradigm Using Working Memory Learning Mechanism 14 Application II: An Autonomous Vision System Based Sensor-Motor Coordination for Open Space Detection 15 Application III: Visual Percepts Learning for Mobile Robot Localization in An Indoor Environment 16 Application IV: An Automatic Natural Scene Recognition Method for Mobile Robot Localization in An Outdoor Environment.