Embodiments related to mapping an environment of a machine-vision system are disclosed. For example, one disclosed method includes acquiring image data resolving one or more reference features of an environment and computing a parameter value based on the image data, wherein the parameter value is responsive to physical deformation of the machine-vision system.