Research on tracking method of flower target

Flowers are a common species in people's lives, but people's understanding of flowers is not comprehensive. The traditional UI display method lacks immersion and interactivity. This paper proposes a method based on the deep learning neural network flower recognition algorithm to recognize the flower target information and replace the traditional UI with the AR information to display the flower information. The tracking and positioning technology used in the floral AR display method is a very important underlying technology. This paper focuses on the tracking technology of flower target, and proposes a server-based flower recognition neural network algorithm recognition algorithm. After identifying the flower information and detection frame, it outputs the detection frame parameters of the flower target, and then combines the OpenCV KCF tracker pair. The method of performing AR display of the flower information has high tracking accuracy and stability.