Multiscale Tracking Method Based on Geometry Estimation

Abstract

The invention discloses a multi-scale tracking method based on geometric estimation, relates to the fields of computer vision and image processing, can perform multi-target tracking, and has fast calculation speed and good real-time performance. The invention includes: input image sequence, track the initial position and size of the target, set the scale and the tracking frame; perform feature extraction and template training on the target in the tracking frame, select the target position and the target scale variable; according to the camera imaging principle, estimate The actual distance between the target and the camera, use the actual distance to estimate the change of the target scale, narrow the value range of the target scale variable; then feed back the scale variable to the first step, repeat the above steps, and finally train to obtain a discriminable target position and scale The filter; the tracking of the scale-changing target is realized through the filter. The invention reduces the calculation amount, ensures the frame rate of the algorithm, improves the real-time performance of the algorithm, and realizes accurate and fast target tracking effect.

Description

technical field [0001] The invention relates to the fields of computer vision and image processing, in particular to a multi-scale tracking method based on geometric estimation. Background technique [0002] With the development of artificial intelligence, computer vision has developed rapidly as an important part of it. Among them, target visual tracking has been widely used in the fields of intelligent transportation, human-computer interaction and intelligent monitoring, and various tracking algorithms have emerged in an endless stream. It has become a domestic and foreign expert. A hot topic of research by scholars. The scale change problem is one of the major problems in the field of target tracking. Many tracking algorithms do not perform well when the target scale changes. As a result, multi-scale tracking algorithms are born, and multi-scale tracking algorithms can establish multiple target scales to adapt to changes in target scales and achieve self-adaptation of t...

AI Technical Summary

Problems solved by technology

[0003] However, the current multi-scale tracking algorithm has a large amount of calculation and poor real-time performance due to the establishment of multiple target scales, and cannot meet the requirements of practical engineering applications. Therefore, the existing technology lacks a fast calculation speed and real-time performance Good multi-object scale tracking method

Images