There are many sources that cause the emergence of geometric deformations in close range images. These deformations are accumulated and not present singly in the image. Therefore, it is necessary to rectify (correct) the image before extracting geometric or semantic data from it. Two methods are available to rectify the close range images. These ones are the parametric and the non-parametric methods. Non-parametric approach does not require knowledge of the parameters of the used camera. Control points and geometric transformations are considered as the two main components in the non-parametric approach. Usually, barrel and perspective deformations are present in close range images. In this paper, we will study the impact of the distribution of control points and the degree of geometric transformation on the correction of the image of these deformations. The test was performed using a close range image of a historical façade. This image was exposed to previous deformations by simulation. The goal is to investigate the effect of the distribution of control points and on the effectiveness of global (linear) and local transformations used to rectify the close range images. It has been demonstrated that the control points located in different parts of the image have different deformation rates, the control points distributed in the center of the image suffers less deformations, and local transformations give the best results when rectifying images with complex deformations.