Estimation of crowd density by counting objects

People counting is a challenging problem in complex situations. The different problems faced are light intensity, perspective transformation, non-human moving object, shadows, occlusion problem, etc. In this paper, we estimate the number of people with the help of background subtraction and scaling. In the first part, we subtract the current image from a background image to estimate the number of people. In the second part, the captured image is brought to the same scale at different position. The main difficulty is to deal with the dynamic changes in the background. Thus, we develop an adaptive background model which will continuously update the changes in the background. These techniques will provide much more accurate solution as well as overcome the drawback with other methods.