AR (augmented reality) algorithm based on real-time tracking

Abstract

The invention relates to the field of image identification and tracking, and provides an AR augmented reality) algorithm based on real-time tracking. The AR algorithm comprises the following steps of: setting a target image; computing a feature code of the target image; setting a maximum tracking point number; obtaining a frame of novel image, and judging whether the novel image has a target tracking point or not, and continuing the tracking if the novel image has the target tracking point; otherwise, computing the target tracking point; computing novel image tracking data; computing a mapping quadrilateral; supplementing a tacking point; updating target tracking image data; and synthesizing an image and displaying the synthesized image on a screen. The AR algorithm is suitable for effective tracking of the target tracking image, so that real-time performance and reliability of tracking are strengthened, and the computing accuracy is improved.

Description

technical field [0001] The invention relates to the field of image recognition and tracking, in particular to realizing real-time tracking of images by using AR technology. Background technique [0002] In recent years, with the advancement of technology, mobile devices such as smartphones and tablets have become indispensable digital entertainment tools for people. AR (augmented reality) technology has been a research hotspot since it was proposed. Augmented reality is a computer system that uses auxiliary information to supplement and enrich the real scene of the user. This system generates information such as graphics, images, sounds, etc. The virtual world and the real world that users feel in reality merge into a world where virtual and real coexist, enabling users to have new visual, auditory and tactile senses. As an important branch of virtual reality technology, augmented reality not only inherits the characteristics of virtual reality (VR), but also highlights it...

AI Technical Summary

Problems solved by technology

AR technology requires the ability to present synthetic scenes in real time, and its implementation is inseparable from scene detection, but robust detection algorithms are often time-consuming, such as the sift algorithm proposed by David G. Lowe and the SURF algorithm proposed by Herbert Bay. Real-time effects cannot be achieved on handheld devices

[0003] Introducing a tracking algorithm with good real-time performance into the AR algorithm can overcome the shortcomings of the detection algorithm, but in practical applications, the accuracy of the tracking algorithm is often affected by the complex environment

Images