A methodology to evaluate/improve the performance of a machine vision system

In this research, the performance of a machine vision system was evaluated and improved by a series of procedures. By considering the hardware component combinations (HCCs) of a machine vision system, a mathematical model was established to represent the relationship between the theoretical and real outputs. By using the mathematical model established for all possible HCCs, an optimal HCC was determined in the evaluation process. Also, the performance of the machine vision system was improved by a second mathematical model established for the optimal HCC. A two-dimensional (2D) machine vision system was used to verify the methodology proposed in this research. After improvement, the accuracy and the repeatability of the machine vision system were ¸-0.001 and ¸±0.002 inch, respectively.