This article addresses the problem of vision-based robot control where the equilibrium is defined via a reference image. Specifically, this work considers the class of general direct solutions that diagonally decouple the control error dynamics also in the general case. Hence, those solutions provide for high versatility and accuracy, whilst the decouplings reduce analysis complexity and can improve system performance. In such a demanding context, the existing general diagonal direct visual servoing technique can present a stability problem arising from its parametrization. This study proposes a new parametrization and its new nonlinear estimator to solve such problem. Theoretical analysis of local stability, convergence, diffeomorphism and of that decoupling property are provided. Experimental results are also obtained using a camera- mounted 6-DoF robotic arm in a challenging scenario.
