Moving Object Detection Based on the Histograms of Oriented Gradients and Cloud Model

We present a Histograms of Oriented Gradients (HOG) features and the cloud model approach for the moving object detection from video sequences. Our model is based on the HOG features for the moving object and then uses the cloud model to find the moving object. First the HOG features are described from the image and then a HOG map is used by the cloud model to detect the moving object. The experiment shows our method is relative effect and has advantage in the detecting moving object.