This paper investigates the global output feedback stabilization problem for a class of nonlinear systems with multiple uncertainties. A distinctive characteristic of the system to be studied is that nonlinearities are bounded by multiplying unknown growth rates with a polynomial form of the output function. The crucial difficulty is how to dominate the nonlinearities effectively on the hypothesis that the output is perturbed by uncertain parameters. An improved dual gain approach is proposed to design the output feedback controller which ensures that all signals of closed-loop systems are globally uniformly bounded and the original system states converge to zero. Finally, the output feedback stabilization problem of a mass-spring mechanical system is used to illustrate the effectiveness of the control strategy.