Since direct manual palpation is not possible in minimally invasive procedures, there is an active field of applied research which aims to retrieve the human sense of touch and feedback tissue properties through artificial tactile feedback. This paper presents an innovative stiffness sensor to be embedded at the tip of a commercial endoscopic camera. The sensor structure is based on multiple cantilever beams, which act as springs with different stiffness when indented into soft tissue. Geometric features mounted on the beams are tracked during physical contact. Movements of thecantilevers result in shape variations of the features in the camera images. The feature size is then segmented and related to the force exerted into the contact location. As beams of different elasticity are integrated, it is possible to estimate the stiffness properties of the soft tissue by employing only visual information. In this paper, Finite Element Analysis (FEA) was implemented to simulate and estimate how contact forces will affect the material and design of the prototype. A calibration device has been developed and used to validate the outcome of the FEA simulations. An experimental test showed the ability of the proposed mechanism to compute the stiffness of a soft phantom.