Weighted Data Fusion Algorithm in the Application of Visual Measurement Network System

Abstract

For the problem on the deviation of actual measurement values between different network positions of visual measurement network system , so present the weighted data fusion algorithm . To distribute the weight coefficients according to the stability of eigenvector and fuse the coordinates , finally get the ultimate expression of fusion estimate. The experiment results present the maximum absolute tolerance of the fusion results is 0.039mm via the weighted data fusion algorithm, the maximum one via traditional data fusion algorithm of the center of point set is 0.056mm , Repeatedly measure 50 times , compared with the traditional algorithm, proposed weighted data fusion algorithm has much better precision and better stability.

A cost-effective planning method for automatic measurement based on task similarity and octopus optimization

“…Calkins [10] developed a unified space measurement network system to transform all the data. Oriented to the visual network system, whose essence is multi-vision measurement, weighted algorithms have been proposed to handle the accuracy problem [11,12]. In addition, multiple optimization methods based on the total least squares have been applied to improve registration uncertainty, which is affected by both measurement errors and configuration errors [13][14][15].…”