In this work the implementation of a computation vision system that classifies blood cells images obtained by means of optical microscopy techniques are described. The classifier is a neural network fed with the feature extracted obtained from images, such as its shape, size and cellular nucleus configuration.