At this step of the algorithm, the background pixels are identified. The values of each pixel are compared using Euclidean distance with the values of the background model. The variance of the pixel is also taken into consideration in the comparison. If the result of the comparison is greater than a threshold, we consider it as foreground and a candidate to be identified. The value of the threshold depends on the global illumination of the scene. It is computed from the average intensity of the pixel and the difference between the darker and lighter pixels.