Calibration performance in merging measurement graph

In this article, we analyze the accuracy of calibration of networked sensors. Networked sensors can be calibrated by maximizing a likelihood of collected measurements. Using the set of the measurements, we can estimate the values of the sensors´ parameters those of objects. These estimated values include estimation errors because of the measurement noises, and we can also estimate the variance of the estimated values. These estimates are computed for one set of sensors and of objects that correspond to the measurements. When one sensor in some such set newly measures one object in another set, then the two sets are merged into one and all estimates of the sensors and the objects may change. In this article, we report an estimation method that can compute the estimates efficiently when some sets of sensors and of objects are merged into one.