Nowadays, there are many methods used to detect the background of an image, with different complexity and different results. Usage varies depending on the application and the necessities. This Project consists in academic software to perform differentiation or segmentation of elements belonging to the background of an image, in order to identify the new objects that appear in the scene. Those objects can be objects that are in constant movement (people, cars, etc.) or objects that reminds in the future (abandoned objects, parked cars, etc.) . In general, background is defined as the objects in the scene that reminds static for long period of time. The environments of these scenes can be classified as internal or external. Each one of them presents substantially different characteristics and some problems that must be resolved. The algorithms implemented in this project will be tested in two types of environment to evaluate their effectiveness and to draw conclusions. We have studied quite different methods, not only in terms of results but also in terms of computational cost and complexity of the algorithms. Automatically arrange from low to high computational cost, the studied algorithms are Frame Difference, Approximate Median Filter, Running Gaussian Average and Mixture of Gaussians finally. In this academic software, the main goal is the easy interaction of the user with the different parameters of each method and the visualization of the effects in the estimation of the background and the foreground. In this memory, is explained briefly the digital image and some aspects of the acquisition of the images are included. The nomenclature used in literature analysis related with this thematic and the different models that have worked are analyzed is also presented. In addition, proceeds to deal in more detail the background removal of an image, the main problems that arise and the specific techniques used. Finally, a review of existing algorithms and their implementation is done.