Please note:

• Information on the updated version of the software is available at the link: CvTracker C++ library.

• Detailed documentation is available at the link: Programmer’s manual.

• Set of cross-platform C++ libraries and examples that allow you to quickly create real-time video processing applications: RapidPixel SDK.

Description

CvTracker (Correlation Video Tracker) C++ library version 8.0.0 is intended for automatic object video tracking. The library is written in C++ (C++17 standard) and uses OpenCV (version 4.5.0 and higher) library to perform forward and backward Fourier transform. The library is compatible with any processors and operating systems supporting C++ compiler (C++17 standard) and OpenCV library (version 4.5.0 and higher). The library provides fast calculation, compatibility with low-power processors, high accuracy and contains a lot of additional functions and modes, which allow using it in camera systems of any configuration. The library contains an advanced tracking algorithm CSRM developed by ConstantRobotics. The library provides tracking of low-contrast and small-sized objects against a complex background.

Performance

Tracking rectangle size 128x128 pixels, grayscale, one video frame processing time:

• 12th Gen Intel Core i5-12500H CPU (one core 2.5 GHz used), search window size 128x128 pixels - 0.5 msec (~2000 fps), search window size 256x256 pixels - 1.4 msec (~714 fps), .

• Jetson Nano (one core 1.43 GHz used), search window size 128x128 pxl - 5.0 msec (~200 fps), search window size 256x256 pixels - 14.5 msec (~68 fps).

• NXP i.MX8M Plus (one core 1.8 GHz used), search window size 128x128 pxl - 5.5 msec (~180 fps), search window size 256x256 pixels - 16.5 msec (~60 fps).

• Raspberry Pi 4 (one core 1.5 GHz used), search window size 128x128 pxl - 4.5 msec (~222 fps), search window size 256x256 pixels - 11.5 msec (~86 fps).

Downloads

• CvTracker C++ library v8.0.0 Programmer’s manual

• Free demo (test) application for version 8.0.0 (for Windows x64): DOWNLOAD

Demo video

Use the Demo Application to estimate the quality of the tracking algorithm.

What’s new in v8.0.0

• The calculation speed has been increased.

• Tracking stability has been improved.

• Added support for YUV, YUYV, UYVY, NV12, RGB, BGR video formats for better tracking stability.

• The number of configurable parameters has been reduced.

• New control commands added.