Collagen fibers and their orientation play a major role in the mechanical behavior of soft biological tissue such as skin. Here, we present a proof-of-principle study correlating mechanical properties with collagen fiber network morphologies. A dedicated multiphoton stretching device allows for mechanical deformations in combination with a simultaneous analysis of its collagen fiber network by second harmonic generation imaging (SHG). The recently introduced Fiber Image Network Evaluation (FINE) algorithm is used to obtain detailed information about the morphology with regard to fiber families in collagen network images. To demonstrate the potential of our method, we investigate an isotropic and an anisotropic ex-vivo dorsal pig skin sample under quasi-static cyclic stretching and relaxation sequences. Families of collagen fibers are found to form a partially aligned collagen network under strain. We find that the relative force uptake is accomplished in two steps. Firstly, fibers align within their fiber families and, secondly, fiber families orient in the direction of force. The maximum alignment of the collagen fiber network is found to be determined by the largest strain. Isotropic and anisotropic samples reveal a different micro structural behavior under repeated deformation leading to a similar force uptake after two stretching cycles. Our method correlates mechanical properties with morphologies in collagen fiber networks.