Enhanced image-based coordinate measurement using a super-resolution method

This paper proposed super-resolution measurement to improve the conventional image-based coordinate measurement system. An undesirable attribute showed by the conventional system is that increasing the accuracy of the system would compromise the field of view of the system. To improve this problem, image super-resolution techniques are proposed to preprocess the observed frames. A Lagrange–Newton method is derived specifically for automatic measurement consideration. Different a priori knowledge was also examined and it was identified that identity model is the most efficient a priori knowledge among the four a priori knowledge tested. Using the Lagrange–Newton method and identity model, an experiment is carried out to evaluate the proposed super-resolution measurement. The result showed that super-resolution measurement increases the accuracy of the system without compromising the field of view. Furthermore, it is also shown that super-resolution measurement can perform measurement on dimension not achievable using the conventional method.