Strain analysis near stress concentration regions is an important tool for structural integrity of mechanical components. However, this analysis becomes more complex when the material starts to deform plastically near the notch root. In this work, a novel experimental-numerical technique for the measurement of the strain distribution on the surface of a deformable body is described, which uses meshless methods and computer vision. The experimental part of this technique is based on the capture of images at different stages of material deformation during a mechanical test. The Scale Invariant Feature Transform (SIFT) is a computer vision algorithm used here to extract distinctive points or features in the captured images. For this purpose, a random texture was painted on the specimen surface. Then, the displacements are experimentally obtained by tracking the positions of successfully matched SIFT points in an undeformeddeformed pair of images. The points provided by SIFT are selected as nodes in a meshless formulation and the moving least square method is used to generate a numerical approximation for the displacement field and its derivates. Thus, the corresponding strain field close to the notch is calculated. To validate the proposed methodology, notched specimens were employed to study the deformation behavior on regions of stress concentration. Experimental results showed good agreement and accuracy when compared to analytical solutions, to simulations by finite elements (ANSYS) and to solutions obtained by using a commercial software based on the digital image correlation technique.