Visual tracking method, visual tracking device, UAV(unmanned aerial vehicle) and terminal device

Abstract

The invention discloses a visual tracking method, a visual tracking device, an UAV(unmanned aerial vehicle) and a terminal device. The method comprises the following steps: obtaining image data of a designated area comprising a tracked object at present and operating parameters of a tracking object; based on the obtained image data of the designated area comprising the tracked object at present, obtaining possible tracked objects in the designated area and generating a tracked object library to select the tracked object; when judging that the tracking object is in relative motion with the selected tracked object, establishing a motion estimation model of the tracked object based on image features of the tracked object in the image data; and based on the motion estimation model, predicatingthe position of the tracked object in the image data of the designated area at the next moment and image features to control the tracking object.

Description

technical field [0001] The invention belongs to the technical field of visual tracking, and in particular relates to a visual tracking method, a visual tracking device, a drone and a terminal device. Background technique [0002] Existing machine autonomous tracking usually includes GPS sensor-based autonomous tracking and vision-based autonomous tracking. Autonomous tracking based on GPS sensors usually requires the body of the tracked object to carry a sensor with GPS positioning function or a similar positioning sensor. The internal processing chip of the tracking machine monitors the position of the tracked object in real time, and tracks the deviation between the current position of the machine and the position of the target tracked object. , so as to guide the tracking machine to follow the tracked object. This tracking method has high limitations and cannot track the tracked object without a sensor with GPS positioning function or a similar positioning sensor. Some v...

AI Technical Summary

Problems solved by technology

Autonomous tracking based on GPS sensors usually requires the tracked object to carry a sensor with GPS positioning function or a similar positioning sensor. The internal processing chip of the tracking machine monitors the position of the tracked object in real time, and tracks the deviation between the current position of the machine and the target tracked object. , so as to guide the tracking machine to follow the tracked object. This tracking method has high limitations and cannot track the tracked object without a sensor with GPS positioning function or a similar positioning sensor.

Some vision-based autonomous tracking methods, such as the visual tracking method of drones, need to actively frame the target object and detect it. Only when the tracked object meets the detection conditions, the drone is driven to perform tracking. This method may It will cause the user to select the tracking object in the image multiple times. Due to human error, the tracked object still cannot meet the detection conditions, resulting in the UAV being unable to perform tracking.

Images