Visual tracking method and device, equipment and storage medium

Abstract

The invention discloses a visual tracking method. The method comprises the following steps: acquiring a to-be-detected image; the to-be-detected image comprises a rectangular target area; determining a pure target area in the rectangular target area; and performing visual tracking by taking the pure target area as a tracking target in a set visual tracking algorithm. According to the visual tracking method disclosed by the invention, the problem that a part of background regions are used as tracking targets in a visual tracking algorithm is solved by determining the pure target region, and the tracking effect of the visual tracking algorithm is improved.

Description

technical field [0001] The present invention relates to the field of computer vision, in particular to a visual tracking method, device, equipment and storage medium. Background technique [0002] With the rapid development of related sciences such as computer network, information storage technology, imaging technology, computer processing power and digital communication, information in the real world can be transformed into digital information that can be processed by computers through computer vision systems. Computer vision uses video sensors to capture real-world images, videos or multi-dimensional data information, and then further uses computers to process these information to realize the functions of human eyes such as target detection, tracking, behavior recognition and analysis. As one of the hottest problems in the world today, computer vision aims to understand the image by detecting the objects contained in the image and defining the relationship between these ob...

AI Technical Summary

Problems solved by technology

[0004] The block tracking strategy does not make full use of the target and its surrounding environment information, so it will affect the tracking effect, specifically the following two points: (1) because the target to be tracked usually has an irregular shape, and the manually marked target bounding box is usually is a regular rectangular frame, then the image blocks sampled by the block tracking algorithm based on the target bounding box may contain too much background area, and using these image blocks for subsequent tracking of the target may cause inaccurate tracking results; (2) In the block tracking strategy, using part of the background information as the apparent representation of the target will affect the tracking effect. What is more serious is that the error will gradually accumulate and propagate during the tracking process, which will affect the accuracy of the tracking algorithm for the target position in subsequent frames. The calculation of , may cause the loss of tracking target

Images