Foreground Object Detection Combining Gaussian Mixture Model and Inter-Frame Difference in the Application of Classroom recording Apparatus
Jun Zhu, Xinhua Zhang · 2018
A new effective approach to detect central coordinate of foreground object in classroom recording application circumstance is proposed in this paper. The new approach includes two steps. The first step is to segment interested blocks from a whole video image by Inter-frame Differences. The second step is to extract the foreground pixels from the interested blocks by Gaussian Mixture Model GMM. The experimental results show that the new algorithm, which combines Gaussian Mixture Model and Inter-frame Differences, performs better than the methods in previous researches in classroom recording application field. The new method is proved to be effective in reducing complexity of calculation with very small expense of accuracy. The adaptability of different number of blocks and different values of block threshold are discussed at the end of the paper.