Underground cross-vision-field target detection tracking method based on multi-template learning

Abstract

The invention discloses an underground cross-view-field target detection tracking method based on multi-template learning, and the method comprises the following steps: S10, carrying out the related filter tracking of a current frame image obtained through the photographing of a current view field through employing an initial template, and obtaining a first response graph, wherein the initial template is a template determined according to a target sample set; s20, performing related filter tracking processing on the current frame image by adopting a process template to obtain a second responsegraph, wherein the process template is a template determined according to a tracking object in each vision field; s30, performing linear weighting processing on the first response graph and the second response graph to obtain a final response graph; and S40, determining the position with the maximum response value in the final response graph as the target position of the tracking target. The method can achieve the long-time tracking of different target objects, and is higher in tracking accuracy.

Description

technical field [0001] The invention relates to the technical field of signal processing, in particular to a multi-template learning-based method for detecting and tracking an underground cross-field target. Background technique [0002] Coal is the basic resource and energy of our country, and an important guarantee for the development of the national economy. Coal mining is one of the key industries in our country. The fully mechanized coal mining face is a high-risk production environment with limited space, high temperature, high humidity, heavy coal ash, low visibility, many high-power equipment and complex working conditions. At present, most of the safety management work for personnel targets in the underground working face adopts the method of manual monitoring, which has the limitations of short duration, narrow coverage and high cost. With the development of intelligent mine construction, the coal mining process gradually tends to be less-manned / unmanned. Therefor...

AI Technical Summary

Problems solved by technology

However, the features extracted by traditional methods are mostly limited to inter-class features, which reduces the accuracy of intra-class matching recognition, and cannot distinguish and identify different targets, which limits the recognition function and leads to low recognition accuracy.

Images