A flexible high-resolution sensor capable of measuring the distribution of both shear and pressure at the plantar interface are needed to study the actual distribution of this force during daily activities, and the role that shear plays in causing plantar ulceration. We have previously developed a novel means of transducing plantar shear and pressure stress via a new microfabricated optical system. However, a force image algorithm is needed to handle the complexity of construction of two-dimensional planar pressure and shear images. Here we have developed a force image algorithm for a micromachined optical bend loss sensor. A neural network is introduced to help identify different load shapes. According to the experimental result, we can conclude that once the neural network has been well trained, it can correctly identify the loading shape. With the neural network, our micromachined optical bend loss Sensor is able to construction the two-dimensional planar force images.