Method for detecting and tracking movement target based on omnidirectional vision of robot

Abstract

A method for detecting and tracking a movement target based on omnidirectional vision of a robot relates to the technical field of robots. The method comprises the following steps: step 1: acquiring and analyzing an image; step 2: detecting the movement target: (2.1), detecting the movement target by adopting an inter-frame differential method, and analyzing an image sequence through an optical flow method after detection; (2.2), comparing two types of detection images in the step (2.1) to obtain incomplete data; (2.3), constructing the incomplete data into complete data by adopting a Gaussian mixture model; detecting the movement target through the Gaussian mixture model; and step 3: tracking the movement target: (3.1), establishing a target reference object, and carrying out target tracking and detection through the reference object, and filtering target color; and (3.2), carrying out positioning processing on the tracked target, and calculating the sequence of a positioned target. By virtue of the method, the target can be conveniently and quickly detected and tracked; and the positioning is accurate and convenient, the efficiency is improved and the safety is high.

Description

Technical field: [0001] The invention relates to the technical field of robots, in particular to a method for detecting and tracking a moving target based on robot omnidirectional vision. Background technique: [0002] 70% of the information in the objective world is obtained by human beings through the eyes, the visual perception organ, and this kind of visual information is mainly obtained by human beings in the form of image information. At present, computer vision technology mainly obtains environmental information through traditional visual sensors. However, traditional visual sensors have a small field of view, and the field of view is mostly within 50°. The local information obtained is limited, resulting in a large blind area of ​​the system, which limits The scope of application of computer vision has been broadened. Omni-directional vision sensors have become a research hotspot worldwide in recent years due to the advantages of one-time acquisition of images in the...

AI Technical Summary

Problems solved by technology

[0005] Due to the complexity of target movement in the environment, there are actually many difficulties in the process of moving target detection and tracking, for example, whether the target trajectory is complex, whether the target has a shape change, whether the color of the background is similar to the color of the target, whether the object in the background is Whether there is motion, whether the light changes are strong, whether the shadows block the moving target and whether the targets block each other, whether the characteristics of the target are easy to select, these will be the difficulties brought by the detection and tracking of the moving target in the multi-video sequence