Accurate measurement of non-common path aberrations (NCPAs) is an important step to be undertaken correctly when operating a real adaptive optics system. NCPAs are defined as the aberrations which are present at the science image but cannot be seen by the wavefront sensor (WFS), basically due to the different placements along the optical path. Compensating these aberrations is required to obtain the best image at the science detector from the resolution point of view. Obtaining the best resolution image, available by a particular optical system having a deformable mirror in the optical path, can be also considered an interesting problem from an abstract point of view, because it can simultaneously compensate for the NCPA and provide the best initial setup for the DM actuators, independent of the accuracy or calibration of the WFS. Phase Diversity (PD) is a very commonly used method of measuring the NCPAs, based in the analysis of two images taken at different focus positions, or just one defocused image. There are also Focal plane Sharpening (FPS) methods, which only deal with the science image, trying to minimize the width of the Point Spread Function by blindly actuating on the DM. The method described in this contribution, Noise Weighted Image Width Minimization (NWIWM), lies in the latter FPS group, and has been developed and tested for the AOLI, EDIFSE and GTCAO projects being developed at Institute of Astrophysics of the Canary Islands (IAC). It is based on the signal to noise analysis of a function describing the width of the PSF with respect to the DM actuations, both zonally and modally, in order to select the actuators or modes to be used during the minimization. A complete description of the algorithm is included, together with simulation results and practical examples obtained within the above mentioned projects.