This one-day course provides hands-on experience with performing computer vision tasks. Examples and exercises demonstrate the use of appropriate MATLAB® and Computer Vision System Toolbox™ functionality.
- Importing, displaying, and annotating videos
- Detecting objects in videos
- Estimating motion of objects
- Tracking a single object or multiple objects
- Removing lens distortion and measuring planar objects
Day 1 of 1
Importing, Visualizing, and Annotating Videos
Objective: Import videos into MATLAB, as well as annotate and visualize them. The focus is on using System Objects™ for performing iterative computations on video frames.
- Importing and displaying video files
- Highlighting objects by drawing markers and shapes like rectangles
- Combining and overlaying two images
- Performing iterative computations on video frames
Objective: Utilize machine learning and deep learning algorithms for complex object detection.
- Marking objects of interest in training images
- Training and using a cascade object detector
- Using a deep learning object detector
Objective: Estimate direction and strength of motion in a video sequence.
- Understanding motion perception in images
- Estimating motion using optical flow methods
Objective: Track single and multiple objects and estimate their trajectory. Handle occlusion by predicting object position.
- Tracking single objects using a Kalman Filter
- Tracking multiple objects using a GNN tracker
Objective: Remove lens distortion from images. Measure size of planar objects.
- Estimating intrinsic, extrinsic, and lens distortion parameters of a camera
- Visualizing the calibration error
- Removing lens distortion
- Measuring planar objects in real-world units