A new method of target tracking and location based on pixel features

Abstract

The invention discloses a novel target tracking and positioning method based on pixel characteristics, which comprises the following steps of: removing human body region: extracting foreground by three frame difference methods for RGB components of pixel points of an image, and binarizing the foreground image to obtain a binary image; detecting the key points of human body in binary image, including hand key points, eyebrow center key points and bipedal midpoint; Marking all connected regions of the binary image; Target tracking and positioning: according to the pixel distribution characteristics of the connected region, the target position and the target positioning point are determined. Motion feature correction: According to the pixel distribution characteristics in the detected connected region, the vector direction of the region is obtained by dimension reduction method and the position of the target point is corrected. The method of the invention can quickly extract motion features, and can quickly and accurately track and locate a moving target in either a simple background or a complex scene, especially when the moving time is short and the target is long and slender, the method is very practical, and has a good popularization prospect.

Description

technical field [0001] The invention belongs to the technical field of machine vision and graphics processing, and in particular relates to a novel target tracking and positioning method based on pixel features. Background technique [0002] Among the traditional moving human body and object tracking algorithms, the tracking algorithm based on color features is better, but this kind of tracking algorithm has poor tracking performance for objects with small targets and light colors, and it is easy to lose the target, so it cannot perform well. Extract features of moving objects. In particular, the traditional method of the present invention has poor tracking and positioning effects for fast-speed, slender-shaped tracking and positioning targets such as golf, badminton, and baseball, and does not have versatility and practicability. Contents of the invention [0003] The technical problem to be solved by the present invention is to overcome the defects of the prior art and ...

AI Technical Summary

Problems solved by technology

[0002] Among the traditional moving human body and object tracking algorithms, the tracking algorithm based on color features is better, but this kind of tracking algorithm has poor tracking performance for objects with small targets and light colors, and it is easy to lose the target, so it cannot perform well. Extract features of moving objects

In particular, the traditional method of the present invention has a poor tracking and positioning effect for fast and slender tracking and positioning targets such as golf, badminton and baseball, and does not have versatility and practicability.

Images