Feature Detection and Extraction

Image registration, interest point detection, extracting feature descriptors, and point feature matching

Local features and their descriptors are the building blocks of many computer vision algorithms. Their applications include image registration, object detection and classification, tracking, and motion estimation. These algorithms use local features to better handle scale changes, rotation, and occlusion. Computer Vision Toolbox™ algorithms include the FAST, Harris, and Shi & Tomasi corner detectors, and the SURF, KAZE, and MSER blob detectors. The toolbox includes the SURF, FREAK, BRISK, LBP, ORB, and HOG descriptors. You can mix and match the detectors and the descriptors depending on the requirements of your application. You can also extract features using a pretrained convolutional neural network which applies techniques from the field of deep learning.

Highlighted Topics

Coordinate Systems

Local Feature Extraction
Detect interest points and extract feature descriptors
Image Registration
Automatic image or video frame registration

Featured Examples

Automatically Detect and Recognize Text in Natural Images

Detect regions in an image that contain text. This is a common task performed on unstructured scenes. Unstructured scenes are images that contain undetermined or random scenarios. For example, you can detect and recognize text automatically from captured video to alert a driver about a road sign. This is different than structured scenes, which contain known scenarios where the position of text is known beforehand.

Open Live Script

Object Detection in a Cluttered Scene Using Point Feature Matching

Detect a particular object in a cluttered scene, given a reference image of the object.

Open Script

Digit Classification Using HOG Features

Classify digits using HOG features and a multiclass SVM classifier.

Open Live Script

Find Image Rotation and Scale Using Automated Feature Matching

Automatically determine the geometric transformation between a pair of images. When one image is distorted relative to another by rotation and scale, use detectSURFFeatures and estimateGeometricTransform to find the rotation angle and scale factor. You can then transform the distorted image to recover the original image.

Open Script

Feature Based Panoramic Image Stitching

Automatically create a panorama using feature based image registration techniques.

Open Script

Video Stabilization Using Point Feature Matching

Stabilize a video that was captured from a jittery platform. One way to stabilize a video is to track a salient feature in the image and use this as an anchor point to cancel out all perturbations relative to it. This procedure, however, must be bootstrapped with knowledge of where such a salient feature lies in the first video frame. In this example, we explore a method of video stabilization that works without any such a priori knowledge. It instead automatically searches for the "background plane" in a video sequence, and uses its observed distortion to correct for camera motion.

Open Live Script

Cell Counting

Use a combination of basic morphological operators and blob analysis to extract information from a video stream. In this case, the example counts the number of E. Coli bacteria in each video frame. Note that the cells are of varying brightness, which makes the task of segmentation more challenging.

Open Live Script

Object Counting

Use morphological operations to count objects in a video stream.

Open Live Script

Object Counting

Use basic morphological operators to extract information from a video stream. In this case, the model counts the number of staples in each video frame. Note that the focus and lighting change in each video frame.

Open Model