Optimal Sensor and Light Source Positioning for Machine Vision

Abstract An optimization approach to automatic sensor and light source positioning for a machine vision task, where geometric measurement and/or object verification is important, is discussed. The goal of the vision task is assumed to be specified in terms of measurements related to edges. The optimal sensor and light source positions are defined in such a way that when the sensor and light source are placed in the optimal positions, we can obtain a picture which produces the minimum variance for the required measurement. Experiments show that the uncertainty in the edge point position is inversely proportional to the contrast across the edge. Using a variant of the Torrance-Sparrow model that takes into account the polarization of the light, the contrasts across edges are computed and used to estimate the variance of the required 2D measurement. An optimization procedure employing mathematical programming techniques uses this information to determine the best positions for the light source and sensor in order to perform the required measurement. A series of experiments was conducted to demonstrate the feasibility of our optimization approach. The optimal positions computed by the program were found to be the best ones in the real experiments. Furthermore, the correlation coefficient between the expected variance and the variance computed from the real pictures was 0.768.