Error analysis of a dimension estimation approach

In this paper, we present an error analysis of a previously reported dimension estimation approach using laser diodes. To assess the reliability of the measurement system, we derive the estimation error with respect to a number of influential variables. We concentrate on the error propagation from the intersection estimation/approximation block to the final dimension estimation block, since the integration of the instrument noise, measurement error, numerical error and processing error can be treated as additive white noise. Our analysis shows that the error minimization problem is equivalent to a feature point position selection problem.