Tracking method based on center detection

Abstract

The invention discloses a tracking method based on center detection. The method comprises the steps of (1) acquiring a video image of the current frame and a video image of the last frame; (2) extracting image background based on the two frames of images obtained from the step (1) and the current tracking speed by means of a displacement frame difference algorithm; (3) obtaining a target image by means of the difference between the obtained images and a background image; (4) determining the coordinates of the center point of the target image; (5) conducting tracking according to the current route if the distance between the center point of the target image and the center of the image of the current frame is smaller than a preset threshold value, and executing the step (6) if not; (6) calculating the field angle of the center point of the target image, and starting tracking according to the deviation between the field angle of the center point of the target image and the field angle of the center point of the image of the current frame to enable the field angle of the center point of the target image to be close to the field angle of the center point of the image of the current frame. The tracking method based on center detection has a good tracking effect.

Description

technical field [0001] The present invention relates to a tracking method, which is used for PTZ tracking of targets meeting certain characteristics, wherein PTZ is the abbreviation of Pan / Tilt / Zoom, which represents pan-tilt omnidirectional (up and down, left and right) movement and lens zooming and zooming control. Background technique [0002] In recent years, with the development of the security industry, various new technologies have emerged on the market and applied in different fields, among which automatic tracking technology is a relatively popular cutting-edge technology. The tracking method can be understood as target detection and tracking, which is a frontier direction that has attracted much attention in the field of computer vision in recent years. It usually includes detecting, recognizing, tracking targets in image sequences of moving targets, and understanding and describing them. [0003] Target analysis includes pattern recognition, image processing, comp...

AI Technical Summary

Problems solved by technology

[0005] There are many aspects involved in target tracking, and the correlation of various aspects and the dependence of processing methods on the front and back are relatively strong, resulting in a relatively slow overall development speed. At present, there is still no good tracking method. The common phenomenon is that tracking is easy. Lost, the tracking effect on fast moving targets is not good, and even the tracking target is shifted during the tracking process

Images