Mobile robot self-localization based on global visual appearance features

The paper presents a novel method for mobile robot localization using visual appearance features. A multidimensional histogram is used to describe the global appearance features of an image such as colors, edge density, gradient magnitude, textures and so on. The matching of histograms determines the location of the robot. The method has been evaluated in an indoor environment, and the system correctly determines the location of 82. 9% of the input scene images.

Authors

• Chao Zhou
• Yucheng Wei
• Tieniu Tan

Keywords

• Mobile robots
• Layout
• Image recognition
• Histograms
• Image segmentation
• Image databases
• Charge-coupled image sensors
• Image processing
• Spatial databases
• Intelligent robots
• Global Features
• Appearance Features
• Mobile Robot
• Visual Appearance Features
• Global Appearance Features
• Indoor Environments
• Scene Images
• Edge Density
• Image Features
• Image Pixels
• Current Position
• Laboratory Environment
• Reference Image
• Current Image
• Matching Results
• Image Database
• Adjacent Positions
• Adjacent Nodes
• Image Histogram
• Node Mapping
• Color Histogram
• Place Recognition
• Image Appearance
• Unstructured Environments
• HSV Color
• Topological Map
• Appearance Information
• Candidate Locations