A scattering medium-based target 4D tracking system and method

Abstract

The invention provides a scattering medium-based target 4D tracking system and method, which solve the technical problem that the conventional target tracking method cannot perform multi-dimension tracking on targets behind blocking objects. The system comprises an LED light source, a lens, a target, a scattering medium, a filter sheet and a detector. The 4D tracking method comprises the steps of:firstly, collecting multiple speckle images, performing denoising preprocessing on the speckle images, calculating self-correlation calculation results of all the denoised speckle images, and obtaining a rotating angle theta of a target according to a correlated coefficient operation algorithm; obtaining a movement result of the target in the optical axis direction z according to self-correlationarea and objective image relations; obtaining movement results of the target in x and y directions according to self-correlation and cross-correlation results. The scattering medium-based target 4D tracking system has the advantages of simple structure and low cost; the tracking method is simple and the tracking precision is high; the system and the method are widely applicable and have a broad application prospect in fields such as biomedicine imaging, military science and civil affairs.

Description

technical field [0001] The invention belongs to the technical field of computational imaging and target tracking, and relates to a 4D optical tracking technology capable of tracking targets behind occluders, specifically a 4D tracking system and method for targets through scattering media, which can be widely used in biological Medical imaging, astronomy, traffic safety and other fields. Background technique [0002] Target tracking technology has always been a hot research topic in the fields of computer vision, biomedicine, astronomy, and earth observation. Target tracking technology can be used to track targets in image sequences. For example, in human-computer interaction design, it is necessary to track human targets feature tracking and identification, in the field of biomedicine, tracking and processing of a certain moving cell in biological tissue, in the field of earth observation, tracking of moving targets such as vehicles and aircraft, etc., according to these ac...

AI Technical Summary

Problems solved by technology

This method is not easily affected by the observation angle, has strong robustness, high accuracy of model matching and tracking, is suitable for various movement changes of maneuvering targets, and has strong anti-interference ability. The update is more complicated and the real-time performance is poor

[0009] The radar tracking method is another target tracking method other than the above-mentioned machine vision. It can realize the continuous measurement and automatic tracking of the distance, azimuth and elevation angle of the target. However, radar tracking mostly uses mechanical scanning to obtain target information. The tracking system is relatively complex, and its working band is easily absorbed and interfered, the real-time tracking performance is poor, and the signal processing algorithm is complex

[0010] The target tracking method and radar tracking method in the above machine vision can track the target, but most of them have the problems of low tracking accuracy, complex tracking system, poor real-time performance, complex calculation, insufficient tracking target information obtained, and narrow application range. And when encountering serious occlusion, none of the above methods can realize target tracking, or the tracking dimension is not high. For the tracking of moving targets, when the target passes through biological tissue or fog, haze and other occlusions, its original visual contour information is almost completely lost, resulting in the inability to track the target, which limits the comprehensive tracking capabilities of existing tracking technologies.

Images