Unmanned aerial vehicle aerial video moving target detection method based on time-space-frequency significance

Abstract

The invention relates to an unmanned aerial vehicle aerial video moving target detection method based on time-space-frequency significance. The method comprises the following steps of extracting the time significance of a video by utilizing a Lucas-Kanade optical flow method; extracting the spatial significance of the image by using color distribution; converting an image from a spatial domain toa frequency domain, extracting frequency domain significance of the image by utilizing a spectral residual method, carrying out linear weighted fusion on time, space and frequency domain significanceto obtain a significance confidence map, binarizing the significance confidence map by setting a threshold value, and extracting a moving target from an aerial video, so that the time domain significance, space domain significance and frequency domain significance are fused, the defects of the domains are overcome through the significance of the other two domains, the detection precision and the detection robustness are improved, the algorithm is simple, and the execution efficiency is high.

Description

technical field [0001] The invention relates to a method for detecting a moving target from an aerial video of a drone, belonging to the field of computer vision. Background technique [0002] UAV aerial video moving target detection is one of the important branches in the field of intelligent analysis of aerial video, and it has extremely important applications in military and civilian fields. At present, experts and scholars at home and abroad have done some research work on the detection of moving objects in aerial video. There is an earlier method based on inter-frame difference. First, the adjacent frames are registered based on feature points or regions, and then the registered adjacent frames are differentiated, and the position of the moving target is judged based on the difference image. However, this method is vulnerable to the accuracy of the registration algorithm. If the registration accuracy is not high, the result of the difference is not accurate enough, wh...

AI Technical Summary

Problems solved by technology

[0004] In order to avoid the deficiencies of the prior art, the present invention proposes a purpose of the present invention is to apply time-space-frequency saliency fusion to the field of unmanned aerial vehicle video moving target detection to solve the problem of low detection accuracy.

Images