Some new filters for image processing are obtained from the wave functions of the two-dimensional quantum oscillator. Such filters are gaussians multiplied by Hermite polynomials and for this reason they will be called Gaussian- Hermite filters. These new quantum filters can be used as smoothing filters and they show good performance when elimination of noise is concerned. Besides of this the new quantum filters can be used to define blurred derivatives and blurred Laplacians for images and in this case the quantum filters are excellent edge detectors. Finally the quantum filters and their derivatives are used to define quantum curvature filters as the Ricci-scalar-curvature filter and the Gaussian-curvature filter. In this last case the quantum filters perform well as curvature detectors and contrast enhancement operators. Our experimental results show that the quantum filters are more efficient than the classical filters and we claim that the quantum image processing will be a very important trend in the near future sensing technology.