Target Detection Algorithm Based on Gaussian Mixture Background Subtraction Model

摘要

Background subtraction method is an effective moving target detection method. The difficulty lies in looking for the ideal, reliable background model for complex scenes and being updated well. While Gaussian mixture model can quickly establish a good background model, process fast, and eliminate the impact of light well. So it becomes one of the commonly used methods in target detection. This paper presents a background subtraction algorithm based on Gaussian mixture. First, background can be obtained accurately using Gaussian mixture model. Then the video lost in the process of establishing is dealt with using background subtraction method. Lastly, detect the target.