Measuring Incompatible Observables by Exploiting Sequential Weak Values