Research on Small Moving Object Detection Based on Improved Gaussian Mixture Model
Sheng Huang, Wei Guiming, Guochun Liu · 2022
As it is difficult to detect small moving targets in wide field of view and long distance environment, this study uses the fluctuation of optical field intensity to enhance the characteristic signal of moving targets, and then combines the local Gaussian mixture model to segment the foreground of the image enhancement area. The detection of small moving targets in different long-distance and complex backgrounds is realized. The result indicates that: The detection algorithm in this study can effectively improve the detection rate of small motion with long distance, wide field of view and low signal-to-noise ratio. The detection rate of drones photographed at 200 meters, 400 meters and 700 meters was 82%.