A Target Tracking Method for Infrared Imaging Based on Centroid Shift of Kernel Function

Abstract

The invention discloses an infrared imaging target tracking method based on kernel function centroid drifting. The method comprises the following steps: 1) inputting infrared images; 2) selecting a corresponding target region in the current frame of infrared images, obtaining a template image and determining global threshold value of the template image; 3) determining a target gray level distribution range according to center grayscale average value of the template image, forming a kernel function, mapping and converting the template image into a quality image, and recording target image total quality and target template binary image; 4) inputting a new frame of infrared image, extracting a wave gate image for comparing at a target prediction position, and calculating a final wave gate position coordinate; 5) mapping the wave gate image obtained after iteration to the quality image and the target template binary image, calculating overlapping degree and quality similarity between the tracking result and template targets, and updating the target template; and 6) skipping to the tracking step 4) until receiving a tracking stop command.

Description

technical field [0001] The invention belongs to the technical field of infrared imaging, in particular to an infrared imaging target tracking method based on kernel function centroid drift. Background technique [0002] In recent years, infrared imaging technology has made great progress and has been widely used in national defense construction and the national economy. In particular, it has become the core technology and important means in military reconnaissance and early warning. Target tracking for infrared imaging is a very important research topic. An effective target tracking algorithm should be able to track the specified target smoothly, and can continue tracking according to the specified rules or stop tracking after a warning signal is issued when the target is occluded or deformed. [0003] Centroid tracking is an important centroid tracking algorithm. It determines the position of the target in the wave gate by calculating the energy moment of the target, and t...

AI Technical Summary

Problems solved by technology

The main disadvantages of the centroid tracking method are as follows: (1) When the tracking target is large and the wave gate cannot completely frame the target, the tracking target cannot be accurately judged by the gray distribution inside and outside the wave gate, such as the tracking target is a bridge, a runway, etc. (2) When the difference between the target gray level and the background gray level is small, the accuracy of the gray level threshold directly affects the stability and accuracy of target tracking; (3) When the target gray level distribution range is large, the gray level The high-level part has a greater impact on the process of calculating the centroid position, and the anti-interference ability is poor; (4) When the target is blocked or a false target appears, it is easy to lose tracking or make mistakes

Images