Student Competition: Computer Vision Training, Matching Image Features

Learn to detect features using a variety of detectors. We will then show how to match features between images.

Features are points or areas of unique content such as corners or blobs.

The Computer Vision Toolbox™ provides the FAST, Harris, and Shi & Tomasi methods for detecting corner features, and the SURF and MSER methods for detecting blob features.

Features are used in two fundamental ways:

• To localize anchor points for use in image stitching, 3-D reconstruction, and stereovision rectification.
• To represent image contents compactly for object detection, recognition or tracking, without requiring image segmentation.

By learning feature detection and matching,, you’ll have the fundamental concepts necessary to perform basic object detection and recognition with features.

We’d like to thank the Cedarville University RoboBoat team for providing a video from their vehicle.
 

Additional Information:

• Local Feature Detection, Extraction, and Matching – Functions, Concepts, and Examples
• showMatchedFeatures function ‘method’ – Display Method property documentation
• Local Feature Detection and Extraction
• Benefits and Applications of Local Features
• Choose a Feature Detector and Descriptor
• Point Feature Types
• Feature Based Panoramic Image Stitching
• Object Detection in a Cluttered Scene Using Point Feature Matching
• Digit Classification Using HOG Features
• Face Detection and Tracking Using the KLT Algorithm
• Image Category Classification Using Bag of Features
• Uncalibrated Stereo Image Rectification
• Stereo Calibration and Scene Reconstruction
• Image Retrieval Using Customized Bag of Features