The VisionLab is a set of components, based on OpenWire technology for advanced computer vision. The components allow rapid development of fully featured motion detection applications for the security industry with almost zero lines of program code.

VisionLab also includes video capturing components based on both Video For Windows (VFW), and DirectShow, so no additional components are needed in order to start building motion detection applications.

The library is fully multithreaded, utilizes well modern multi-core systems, and runs at full native performance.

Also includes Visual Graphical Editor for Codeless Development and Visual Live Bindings.

VisionLab contains:

• Video Capture - captures analog or digital video sources like FireWire (IEEE1394) cameras, DV camcorders, USB web cams, IP cameras, TV cards, PCI composite cards, USB devices, with or without audio.
• Video Recorders - Records in compressed or uncompressed AVI.
• Video Players - Play compressed or uncompressed AVI, WMV files or IP streams.
• Image Generators - images from bitmaps.
• Video Layers - Applies video layers over the video frames.
• Timing components - Clocks and Watchdog Timers.
• Video Filters - Noise reduction filters, crop, and resize components.
• Motion Detector - Detects motion in the video.
• Adaptive threshold - Performs Adaptive threshold algorithm for edge detection.
• Canny - Performs Canny algorithm for edge detection.
• Hough Lines - Finds lines in the image using Canny edge detection and Hough transform.
• Probable Lines - Finds probable line segments in the image using Canny edge detection and Hough transform.
• Hough Circles - Finds circles in the image using Hough transform.
• Contour detection - Finds contours in binary image.
• Find Targets - Finds possible targets in the image. The targets are detected as strong corners in the image. The targets then can be tracked by the Track Targets component.
• Track Targets - Tracks moving targets on the image. The targets are tracked as strong corners in the image.
• Extract Robust Features - Extracts robust features using SURF algorithm.
• Haar Detect Objects - Detects objects by using Haar like features.
• HOG Detect Objects - Detects objects by using Histogram of Oriented Gradients.
• HOG Training - Prepares Histogram of Oriented Gradients training.
• Label Connected Components - Labels connected components in the image.

Typical applications:

• Motion detection - Create security systems with motion detection, or activate the video recording when motion is detected.
• Detect objects - Detect objects with different shapes in the video.
• Face Detect and recognition - Detect and/or recognize faces in images or video.
• Fingerprint compare - Process and compare fingerprint in image or video.
• Find and Track Targets - Find and track objects in the video.