The video capture class allows you to process video files into individual frames. You can then use different operators/functions such as read, get, and set to retrieve or manipulate the frames.

Video Capture

Camera calibration is an important pre-processing step in computer vision that corrects for the radial and tangential distortion introduced by a camera's lens. This is done by estimating the camera's intrinsic parameters (e.g., focal length, optical center) and extrinsic parameters (rotation and translation relative to the scene) from a set of calibration images. OpenCV provides functions such as `calibrateCamera`, `getOptimalNewCameraMatrix`, and `undistort` to compute these parameters and remove the distortion.

Camera Calibration

An affine transformation represents a relationship between two images through rotation, translation, and scaling. OpenCV has a few functions such as affineTransform2D. affineTransform3D which create 2x3 matrix transformation matrices between two coordinates systems in the same dimension.

Affine Transform

OpenCV provides several object tracking algorithms, including the BOOSTING, MIL, Kernelized Correlation Filter, Discriminative Correlation Filter, MedianFlow, TLD, MOSSE, and GOTURN trackers.

Object Tracking in OpenCV

An important pre-processing step in computer vision is converting to different color schemes such as gray or HSV. OpenCV makes the seamless using cv2.cvtColor where you can input you image and use flags such as COLOR_BGR2RGB to convert it between color schemes.

Converting image in OpenCV

Smoothing and blurring is image processing step that reduces noise and detail in an image. It is important for edge detection and thresholding. OpenCV has few different options such as:
- Average blur
- Guassian blur
- Median filtering
- Bilateral filtering

Smoothing & Blurring in OpenCV

In computer vision, feature matching involves finding identical features across different images. OpenCV provides a brute-force matcher technique for this task, which takes a feature descriptor from one image and matches it with others using a distance calculation, which defaults to the L2 norm.

Feature Matching in OpenCV

OpenCV implements the Lucas-Kanade sparse optical flow method via the `cv2.calcOpticalFlowPyrLK()` function. This method performs a least-squares fit on the motion of a 3x3 pixel neighborhood (9 points) around each target point, assuming small movement between consecutive frames to satisfy the brightness constancy constraint. The function accepts the previous frame, the next frame, and the tracking points from the previous frame, returning the updated points and a status vector indicating successful tracking.

Lucas-Kanade Optical Flow in OpenCV

OpenCV is an open-source computer vision library that provides documented classes and functions for tasks such as face detection, object identification, three-dimensional model extraction, motion tracking, frame stitching, camera calibration, image processing, and feature matching. It can be used with C++, Python, Java, and MATLAB.

University of Michigan - Ann Arbor

Google

As computer vision methods have become relatively standardized, it is efficient to utilize established software libraries and tools. Prominent examples include OpenCV, the SciPy ndimage Module, TensorFlow, PyTorch Vision, and the MATLAB Vision Toolbox.

Computer Vision Libraries

OpenCV

MATLAB Vision Toolbox provides algorithms, functions, and apps for designing and testing computer vision, 3D vision, and video-processing systems. It also supports training deep-learning and machine-learning models and using multicore or GPU systems.

MATLAB Vision Toolbox

The `scipy.ndimage` module provides functions for multidimensional image processing in Python. It includes operations for filtering, mathematical morphology, interpolation, object measurements, and feature extraction on NumPy arrays that represent n-dimensional images.

Learn Before

Related

Learn After