In Dynamical Astronomy, the data used to fit theories are positional observations. If the data are reduced using computers, the measure itself is not, and informations are lost. The classical way is to use photographic plates measured by hand. The use of automatic machine permits to save time but not to increase the amount of data. The arrival of silicium targets and CCD allows to acquire much more data directly stored on computer. We used this technique associated with array processors. So we succeeded to increase the accuracy of the data. This method was applied to the observation of the motion of some asteroids and of the Galilean satellites of Jupiter. © 1989 Kluwer Academic Publishers.