Multi-target continuous tracking system and method based on cloud deck control technology

Abstract

The invention relates to a multi-target continuous tracking system and a method based on a cloud deck control technology. The system comprises a main control board, a bolt cloud deck assembly and a ball cloud deck assembly, wherein the bolt cloud deck assembly and the ball cloud deck assembly are connected with the main control board; the bolt cloud deck assembly comprises a wide-angled camera, abolt motor for driving the wide-angled camera to move, and a bolt cloud deck control module for controlling the bolt motor to rotate; the ball cloud deck assembly comprises an integral zooming machinecore, a ball motor for driving the integral zooming machine core to move, and a ball cloud deck control module for controlling the ball motor to rotate; the main control plate comprises a main control module, a GPS (Global Position System) module, a communication module and an image processing module; the GPS module, the communication module and the image processing module are connected with themain control module; the wide-angled camera and the integral zooming machine core are connected with the image processing module; the bolt cloud deck control module and the ball cloud deck control module are connected with the main control module; the main control plate is connected with an upper computer. The system has the advantages that through the multi-camera linkage, the real-time monitoring is realized; the prevention effect is achieved.

Description

technical field [0001] The invention relates to the field of security technology, in particular to a multi-target continuous tracking system and method based on pan-tilt control technology. Background technique [0002] At present, the continuous tracking of targets based on multiple cameras mostly uses the feature data of the monitored background and targets to be transmitted to the background server, and the background server uses the SIFT (Scale Invariant Feature Transform) feature matching of the background image to detect the repeated areas between the cameras. According to the key points, the corresponding relationship between the overlapping areas between the cameras is established, and then the tracked target is cut out to calculate the SIFT feature value, which is matched with the target of the adjacent camera, so as to complete the continuous tracking of the target. [0003] The tracking target in the prior art is subject to factors such as illumination, weather, o...

AI Technical Summary

Problems solved by technology

[0003] The tracking target in the existing technology is subject to factors such as illumination, weather, occlusion, and similarity in clothing, which lead to uncertainty in target matching. In addition, it is impossible to judge whether it is a valid target from the actual calculated trajectory and velocity.

When the target is mixed into the background of the same group, for example, when the target is dressed similarly to other targets in the background, it is easy to follow or track wrongly

With the method of multi-camera splicing, the target cannot be zoomed in, and more pixel features cannot be obtained.

The back-end server load is directly affected by the number of front-end cameras, so it cannot be easily expanded and extended

[0004] Now high-point monitoring, the linkage between monitors is still in its infancy. How to transmit the coordinate information o

Images