Improved Adaptive Gaussian Mixture Model Moving Object Detection Method

Abstract

The invention provides an improved adaptive Gaussian mixture model moving target detection method. Gauss distributions and parameters are initialized, and k Gauss distributions are sorted according to weights omega<i, t>, wherein the i belongs to [1, K]. The distributions are updated according to whether the distributions fit the existing Gauss distributions, wherein the high-weight distributions are updated, the distributions with close weights are merged, the low-weight distributions are deleted, and new distributions are added. The method updates values of the learning rate and model distribution number and solves data redundancy. Meanwhile, parameter values are updated and can be better adapted to environment change. The method is a moving target detection method based on computer video image processing technology, can be applied to real-time video processing systems, provides key theories and application foundation for vehicle testing or fire monitoring, and has good application prospect.

Description

technical field [0001] The invention relates to an improved adaptive Gaussian mixture model moving target detection method. Background technique [0002] At present, optical flow method, frame difference method, and background subtraction method are commonly used algorithms for moving object detection. Among them, the optical flow method is computationally complex, requires specialized hardware support, and has poor real-time performance and practicability. Although the frame difference method can effectively remove the static background, the extracted target is often rough and larger than the actual moving target outline, and there will be holes and "double shadows" in the target. Background subtraction can detect moving objects in changing environments, but requires real-time updates to the background image. [0003] From a practical point of view, background subtraction is the most widely used method for moving object detection. Among them, the most commonly used backg...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide an improved adaptive Gaussian mixture model moving target detection method to solve the problems existing in the prior art because the values ​​of its learning rate and model distribution number are relatively fixed, which easily causes data redundancy; at the same time, relatively fixed There are still problems such as not being able to better adapt to environmental changes in the parameter values

Images